Update app.py

app.py

@@ -1,1292 +1,265 @@
-# from __future__ import annotations
-# import os
-# import re
-# import uuid
-# import json
-# import time
-# import atexit
-# import fitz  # PyMuPDF
-# import docx
-# import markdown
-# from datetime import datetime, timezone
-# from typing import List, Dict, Any, Optional
-
-# import gradio as gr
-# import numpy as np
-# from openai import OpenAI
-# import weaviate
-# from weaviate.classes.init import Auth
-# from weaviate.classes.config import Configure, Property, DataType
-# from weaviate.classes.query import Filter, MetadataQuery
-
-# # -------------------- Configuration --------------------
-# MODEL_NAME = os.getenv("MODEL_NAME", "openai/gpt-oss-120b")
-# DEEPINFRA_API_KEY = os.getenv("DEEPINFRA_API_KEY", "kPEm10rrnxXrCf0TuB6Xcd7Y7lp3YgKa")
-# BASE_URL = os.getenv("BASE_URL", "https://api.deepinfra.com/v1/openai")
-
-# WEAVIATE_URL = os.getenv("WEAVIATE_URL", "htorgbgpt4w63nvf1yeuw.c0.us-west3.gcp.weaviate.cloud")
-# WEAVIATE_API_KEY = os.getenv("WEAVIATE_API_KEY", "ZUd6clB5WmYzVGkxeU40cl96NTY5UkViUlVzY05Md3IzQ0JKelBZQmxGZHRPeGpCeGdxS1FUNnlYUkFFPV92MjAw")
-
-# MEMORY_FILE = os.getenv("MEMORY_FILE", "chat_memory.json")
-# LOG_FILE = os.getenv("LOG_FILE", "interaction_logs.json")
-
-# # Check for missing environment variables
-# if not DEEPINFRA_API_KEY or not WEAVIATE_URL or not WEAVIATE_API_KEY:
-#     raise ValueError("One or more required environment variables are not set: DEEPINFRA_API_KEY, WEAVIATE_URL, WEAVIATE_API_KEY")
-
-# # -------------------- Clients --------------------
-# llm_client = OpenAI(api_key=DEEPINFRA_API_KEY, base_url=BASE_URL)
-
-# weaviate_client = weaviate.connect_to_weaviate_cloud(
-#     cluster_url=WEAVIATE_URL,
-#     auth_credentials=Auth.api_key(WEAVIATE_API_KEY),
-# )
-
-# # -------------------- Helpers --------------------
-# UTC = timezone.utc
-
-# SYSTEM_PROMPT_BASE = (
-#     "You are **EduNatives Assistant**, a helpful, friendly, and precise academic/career guide. "
-#     "You serve three primary audiences: Students, Universities/Researchers, and Companies.\n\n"
-#     "Goals by audience:\n"
-#     "- Students: registration/account help; finding internships/scholarships; connecting with mentors or professors; querying projects; applying for jobs/opportunities; joining project teams.\n"
-#     "- Universities/Researchers: publish research or announce events; connect/collaborate with students.\n"
-#     "- Companies: post jobs/internships/graduate roles; discover student talent.\n\n"
-#     "General rules:\n"
-#     "- Reply in the user's language (Arabic if the user writes Arabic; otherwise English).\n"
-#     "- Use data from Weaviate collections (Job, Opportunities, Project) when relevant.\n"
-#     "- Be concise, step-by-step, and action-oriented (lists, bullets, checklists).\n"
-#     "- If information is unavailable, state that clearly and suggest the next best step.\n"
-#     "- For CV analysis, extract skills/experience and recommend matching opportunities.\n"
-#     "- Assist with applications and team matching.\n"
-#     "- Ensure that all generated prompts are phrased using positive reinforcement."
-# )
-
-
-# def get_rfc3339_time() -> str:
-#     """Return current UTC time in RFC3339 format"""
-#     return datetime.now(UTC).isoformat().replace('+00:00', 'Z')
-
-# # --- CV Skills Extraction (Regex baseline, can replace with NLP model later) ---
-# _SKILL_REGEX = re.compile(r"\b(Natural Language Processing|Building Information Modeling|Search Engine Optimization|Search Engine Marketing|Aerospace Engineering & Management|Computational Fluid Dynamics|Kotlin Multiplatform|Google Cloud Platform|Social Media Marketing|Aerospace Engineering|Microsoft SQL Server|Amazon Web Services|Finite Element Analysis|Technology-based Management|Autodesk Inventor|Emotional Intelligence|Aerospace Engineering & Operations Management|Content Marketing|Presentation Skills|Interpersonal Skills|Critical Thinking|Financial Modeling|Decision Making|Process Improvement|Time Management|Lean Manufacturing|Project Management|Microsoft Excel|Data Visualization|Computer Vision|Machine Learning|Deep Learning|Attention to Detail|Six Sigma|Risk Analysis|Data Analysis|Data Science|Communication|Collaboration|Teamwork|Leadership|Management|Adaptability|Creativity|Innovation|Negotiation|Android|Angular|Ansible|Apache|ArcGIS|Arduino|Asana|ASP\.NET|AutoCAD|Azure|Bash|BIM|Business Analysis|C\+\+|C#|CAM|Cassandra|CATIA|CentOS|Chef|CI/CD|Civil 3D|Content Marketing|CRM|CSS|Data Mining|Django|Docker|Elasticsearch|Email Marketing|ERP|ETABS|ETL|Express\.js|Facebook Ads|Firebase|Flask|Flutter|FPGA|Fusion 360|GCP|Git|GitHub|GitLab|Go|Google Ads|Google Analytics|GraphQL|Hadoop|HTML|HubSpot|iOS|Java|JavaScript|Jenkins|Jira|Jupyter Notebook|Kanban|Keras|Kotlin|Kubernetes|LabVIEW|Laravel|LESS|Linux|Lua|macOS|Marketo|MATLAB|Matplotlib|MongoDB|Multisim|MySQL|Nginx|NLP|Node\.js|NoSQL|Nuxt\.js|NumPy|Next\.js|Objective-C|Oracle Database|Oracle|OrCAD|Pandas|Perl|PHP|PLC|Plotly|PostgreSQL|Power BI|PowerShell|Problem Solving|Puppet|PSpice|Python|PyTorch|Raspberry Pi|React Native|React|Red Hat|Redis|Revit|Ruby on Rails|Ruby|Rust|Salesforce|SAP2000|SAP|Sass|SCADA|Scala|Scikit-learn|Scrum|Seaborn|SEM|SEO|Simulink|SketchUp|Slack|SolidWorks|Spring Boot|SQL|SQLAlchemy|SwiftUI|Swift|Tableau|Terraform|TensorFlow|Trello|TypeScript|Ubuntu|Verilog|VHDL|Vue\.js|Waterfall|Windows|WordPress|Xamarin|Analytical Skills)\b", re.IGNORECASE)
-
-# def extract_skills_from_text(cv_text: str) -> List[str]:
-#     skills = list({m.group(0).lower() for m in _SKILL_REGEX.finditer(cv_text)})
-#     return [s.capitalize() for s in skills]
-
-
-# # --- Process uploaded file (PDF, DOCX, TXT) ---
-# def process_uploaded_file(file_obj: Any) -> dict | None:
-#     if not file_obj:
-#         return None
-#     file_path = file_obj.name
-#     filename = os.path.basename(file_path)
-#     text_content = ""
-#     try:
-#         if filename.lower().endswith(".pdf"):
-#             with fitz.open(file_path) as doc:
-#                 for page in doc:
-#                     text_content += page.get_text()
-#         elif filename.lower().endswith(".docx"):
-#             docp = docx.Document(file_path)
-#             for p in docp.paragraphs:
-#                 text_content += p.text + "\n"
-#         elif filename.lower().endswith(".txt"):
-#             with open(file_path, "r", encoding="utf-8") as f:
-#                 text_content = f.read()
-#         else:
-#             return {"error": f"Unsupported file type: {filename}"}
-
-#         skills = extract_skills_from_text(text_content)
-#         return {"content": text_content.strip(), "skills": skills, "filename": filename}
-
-#     except Exception as e:
-#         return {"error": f"Error processing file {filename}: {e}"}
-
-# # --- Chat history HTML formatter (for Gradio) ---
-# def format_chat_html(history: List[Dict[str, str]]) -> str:
-#     html = "<div class='chatbot'>"
-#     for msg in history:
-#         role = msg["role"]
-#         content = msg["content"]
-#         if role == "user":
-#             html += f"<div class='user-bubble'>{content}</div>"
-#         else:
-#             # Render bot response as Markdown
-#             html_content = markdown.markdown(content, extensions=['tables'])
-#             html += f"<div class='bot-bubble'>{html_content}</div>"
-#     html += "</div>"
-#     return html
-
-# # ================================
-# # Part 2 — Weaviate Collections + Query + RAG
-# # ================================
-
-# # -------------------- Ensure collections --------------------
-# def ensure_collections():
-#     # Application collection (no vectorizer)
-#     if not weaviate_client.collections.exists("Application"):
-#         weaviate_client.collections.create(
-#             name="Application",
-#             properties=[
-#                 Property(name="applicationId", data_type=DataType.TEXT),
-#                 Property(name="jobId", data_type=DataType.TEXT),
-#                 Property(name="applicantName", data_type=DataType.TEXT),
-#                 Property(name="applicantEmail", data_type=DataType.TEXT),
-#                 Property(name="coverLetter", data_type=DataType.TEXT),
-#                 Property(name="cvText", data_type=DataType.TEXT),
-#                 Property(name="skills", data_type=DataType.TEXT_ARRAY),
-#                 Property(name="createdAt", data_type=DataType.DATE),
-#             ],
-#             vectorizer_config=Configure.Vectorizer.none()
-#         )
-
-#     # Team collection (no vectorizer)
-#     if not weaviate_client.collections.exists("Team"):
-#         weaviate_client.collections.create(
-#             name="Team",
-#             properties=[
-#                 Property(name="teamId", data_type=DataType.TEXT),
-#                 Property(name="name", data_type=DataType.TEXT),
-#                 Property(name="projectId", data_type=DataType.TEXT, skip_vectorization=True),
-#                 Property(name="members", data_type=DataType.TEXT_ARRAY),
-#                 Property(name="skills", data_type=DataType.TEXT_ARRAY),
-#                 Property(name="createdAt", data_type=DataType.DATE),
-#                 Property(name="creatorId", data_type=DataType.TEXT),
-#                 Property(name="idea", data_type=DataType.TEXT),
-#             ],
-#             vectorizer_config=Configure.Vectorizer.none()
-#         )
-    
-#     # RAG Collections (with vectorizer)
-#     rag_collections = {
-#         "Job": [
-#             Property(name="jobId", data_type=DataType.TEXT), Property(name="title", data_type=DataType.TEXT),
-#             Property(name="companyName", data_type=DataType.TEXT), Property(name="description", data_type=DataType.TEXT),
-#             Property(name="skills", data_type=DataType.TEXT_ARRAY), Property(name="workplaceType", data_type=DataType.TEXT),
-#             Property(name="employmentType", data_type=DataType.TEXT_ARRAY), Property(name="requirements", data_type=DataType.TEXT),
-#         ],
-#         "Opportunities": [
-#             Property(name="opportunityId", data_type=DataType.TEXT), Property(name="title", data_type=DataType.TEXT),
-#             Property(name="organization", data_type=DataType.TEXT), Property(name="description", data_type=DataType.TEXT),
-#             Property(name="type", data_type=DataType.TEXT), Property(name="skills", data_type=DataType.TEXT_ARRAY),
-#         ],
-#         "Project": [
-#             Property(name="projectId", data_type=DataType.TEXT), Property(name="title", data_type=DataType.TEXT),
-#             Property(name="description", data_type=DataType.TEXT), Property(name="skills", data_type=DataType.TEXT_ARRAY),
-#             Property(name="creatorName", data_type=DataType.TEXT),
-#         ]
-#     }
-
-#     for name, properties in rag_collections.items():
-#         if not weaviate_client.collections.exists(name):
-#             weaviate_client.collections.create(
-#                 name=name,
-#                 properties=properties,
-#                 vectorizer_config=Configure.Vectorizer.text2vec_openai()
-#             )
-
-# ensure_collections()
-
-# # -------------------- Query Weaviate --------------------
-# def query_weaviate_collection(class_name: str, query_text: str, limit: int = 5) -> List[dict]:
-#     """Performs a hybrid search on a Weaviate collection."""
-#     try:
-#         collection = weaviate_client.collections.get(class_name)
-#         response = collection.query.hybrid(
-#             query=query_text,
-#             limit=limit,
-#             query_properties=["title^2", "description", "skills"] if class_name == "Job" else None
-#         )
-#         return [obj.properties for obj in response.objects]
-#     except Exception as e:
-#         print(f"[Weaviate Query Error] {e}")
-#         return []
-
-# # -------------------- RAG Prompt Builder --------------------
-# def build_rag_prompt(user_question: str, retrieved_items: List[dict], class_name: str) -> str:
-#     if not retrieved_items:
-#         return ""
-#     context_parts = [json.dumps(item, indent=2, ensure_ascii=False) for item in retrieved_items]
-#     context_block = "\n\n---\n\n".join(context_parts)
-
-#     return f"""
-#     User Question: "{user_question}"
-    
-#     You are an expert AI assistant. Your ONLY source of information is the structured JSON data provided below under "Retrieved Data". If the data section is empty, you must state that no results were found.
-    
-#     **Instructions:**
-#     1.  Analyze all JSON objects provided.
-#     2.  Convert the data for each item into a clear, readable summary. Use Markdown for formatting (headings, lists).
-#     3.  Present all important details like title, company/creator, description, and skills.
-    
-#     Retrieved Data from '{class_name}' collection:
-#     {context_block}
-#     """
-
-# # -------------------- RAG Answer --------------------
-# def rag_answer(user_question: str, class_name: str, top_k: int = 3) -> (str, List[dict]):
-#     retrieved = query_weaviate_collection(class_name, user_question, limit=top_k)
-#     if not retrieved:
-#         return f"Sorry, I couldn't find any results related to '{user_question}' in the {class_name} database.", []
-
-#     prompt = build_rag_prompt(user_question, retrieved, class_name)
-
-#     try:
-#         resp = llm_client.chat.completions.create(
-#             model=MODEL_NAME,
-#             messages=[
-#                 {"role": "system", "content": SYSTEM_PROMPT_BASE},
-#                 {"role": "user", "content": prompt}
-#             ],
-#             temperature=0.3, max_tokens=4096
-#         )
-#         answer = resp.choices[0].message.content or ""
-#     except Exception as e:
-#         print(f"[RAG LLM Error] {e}")
-#         answer = "⚠️ Sorry, I couldn't process that. Try again later."
-
-#     return answer, retrieved
-# # ================================
-# # Part 3 — Conversation State Machine + Embedding Recommendations
-# # ================================
-
-# # -------------------- Embedding helpers --------------------
-# def compute_embedding(text: str) -> List[float]:
-#     """Compute embedding using the OpenAI-compatible client."""
-#     try:
-#         resp = llm_client.embeddings.create(model="Qwen/Qwen3-Embedding-8B", input=text, encoding_format="float")
-#         if hasattr(resp, "data") and resp.data:
-#             return resp.data[0].embedding
-#     except Exception as e:
-#         print("[compute_embedding] error:", e)
-#     return []
-
-# def cosine_similarity(a: List[float], b: List[float]) -> float:
-#     if not a or not b: return 0.0
-#     va = np.array(a, dtype=float)
-#     vb = np.array(b, dtype=float)
-#     denom = (np.linalg.norm(va) * np.linalg.norm(vb))
-#     return float(np.dot(va, vb) / denom) if denom != 0 else 0.0
-
-# # -------------------- Recommendations by embedding --------------------
-# def recommend_jobs_by_embedding(cv_text: str, top_k: int = 5, jobs_fetch_limit: int = 200) -> str:
-#     """
-#     NOTE: This function computes embeddings on-the-fly for jobs, which is slow for large datasets.
-#     For production, pre-calculate and store job embeddings in Weaviate.
-#     """
-#     skills = extract_skills_from_text(cv_text or "")
-#     user_text = " ".join(skills) if skills else (cv_text or "")[:500]
-#     user_emb = compute_embedding(user_text)
-#     if not user_emb:
-#         return "⚠️ Unable to compute embedding for your CV. Try again or check API keys."
-
-#     try:
-#         jobs_col = weaviate_client.collections.get("Job")
-#         fetched = jobs_col.query.fetch_objects(limit=jobs_fetch_limit)
-#         if not fetched.objects: return "⚠️ No jobs found in the database."
-#     except Exception as e:
-#         print("[recommend_jobs_by_embedding] Weaviate fetch error:", e)
-#         return "⚠️ Could not fetch jobs from the database."
-
-#     scored_jobs = []
-#     for obj in fetched.objects:
-#         props = obj.properties
-#         job_text_parts = [*(props.get("skills") or []), props.get("title", ""), (props.get("description") or "")[:1000]]
-#         job_text = " ".join(filter(None, job_text_parts)).strip()
-#         if not job_text: continue
-        
-#         job_emb = compute_embedding(job_text)
-#         if job_emb:
-#             score = cosine_similarity(user_emb, job_emb)
-#             scored_jobs.append((score, props))
-
-#     if not scored_jobs:
-#         return "⚠️ No jobs could be compared. The database might be empty or embeddings could not be generated."
-
-#     scored_jobs.sort(key=lambda x: x[0], reverse=True)
-#     top = scored_jobs[:top_k]
-
-#     lines = [f"### Here are the top {len(top)} recommended jobs for you:"]
-#     for score, props in top:
-#         lines.append(
-#             f"**{props.get('title', 'N/A')}** at *{props.get('companyName', 'N/A')}*\n"
-#             f"- **Match Score**: {score:.2%}\n"
-#             f"- **Skills**: {', '.join(props.get('skills', []))}\n"
-#             f"---"
-#         )
-#     return "\n".join(lines)
-    
-# # -------------------- INTENT ROUTING & KNOWLEDGE BASE --------------------
-# def route_intent(text: str) -> Optional[str]:
-#     low = text.lower()
-#     # KB intents
-#     if any(k in low for k in ["register", "تسجيل"]): return "kb_registration"
-#     if ("who" in low and "edunatives" in low) or ("مين" in low and "edunatives" in low): return "kb_about"
-#     if ("how" in low and "apply" in low) or ("كيف" in low and "وظيفة" in low): return "kb_apply_job"
-#     if ("contact" in low and "support" in low) or ("اتواصل" in low and "دعم" in low): return "kb_contact_support"
-#     # RAG intents
-#     if any(k in low for k in ["job", "وظيفة", "internship"]): return "rag_job"
-#     if any(k in low for k in ["project", "مشروع"]): return "rag_project"
-#     if any(k in low for k in ["opportunity", "فرصة", "scholarship"]): return "rag_opportunities"
-#     return None
-
-# def kb_fallback(intent: str) -> Optional[str]:
-#     kb = {
-#         "kb_registration": "To register as a student, please visit our website, click on 'Register', and fill out the form. [Arabic: للتسجيل كطالب، قم بزيارة موقعنا، اضغط على 'تسجيل'، واملأ البيانات المطلوبة.]",
-#         "kb_about": "EduNatives is an academic and career platform connecting students with opportunities, projects, and mentoring from universities and companies. [Arabic: EduNatives هي منصة أكاديمية ومهنية تربط الطلاب بالفرص والمشاريع والإرشاد من الجامعات والشركات.]",
-#         "kb_apply_job": "You can apply for a job by finding it on the platform, clicking 'Apply', and following the steps, which include uploading your CV. [Arabic: يمكنك التقديم على وظيفة بالبحث عنها في المنصة، الضغط على 'قدم الآن'، واتباع الخطوات التي تتضمن تحميل سيرتك الذاتية.]",
-#         "kb_contact_support": "You can contact support via email at support@edunatives.com. [Arabic: يمكنك التواصل مع الدعم عبر البريد الإلكتروني support@edunatives.com.]"
-#     }
-#     return kb.get(intent)
-    
-# # -------------------- Conversation Session helpers --------------------
-# def initial_session() -> dict:
-#     return {"state": "idle", "data": {}}
-
-# def handle_uploaded_cv_for_session(session: dict, uploaded_file: Any) -> (str, dict):
-#     if not uploaded_file: return "⚠️ No file received.", session
-#     doc_info = process_uploaded_file(uploaded_file)
-#     if not doc_info or "error" in doc_info:
-#         return f"⚠️ Error processing CV: {doc_info.get('error', 'unknown error')}", session
-
-#     session["data"]["cvText"] = doc_info.get("content", "")
-#     session["data"]["cvSkills"] = doc_info.get("skills", [])
-    
-#     st = session.get("state")
-#     if st == "apply_wait_cv":
-#         session["state"] = "apply_jobtitle"
-#         detected = ", ".join(session["data"]["cvSkills"]) if session["data"]["cvSkills"] else "none"
-#         return f"CV received. Detected skills: {detected}. Which job title are you applying for?", session
-#     if st == "recommend_wait_cv":
-#         rec_text = recommend_jobs_by_embedding(session["data"]["cvText"])
-#         return f"Here are job recommendations based on your CV:\n\n{rec_text}", initial_session()
-
-#     return "CV uploaded. What's next?", session
-
-# # =================================================================
-# # FIXED: Functions to save data to Weaviate
-# # =================================================================
-
-# def save_application_to_weaviate(app_data: dict) -> bool:
-#     """Saves a new application object to the Weaviate 'Application' collection."""
-#     try:
-#         applications = weaviate_client.collections.get("Application")
-#         applications.data.insert(properties=app_data)
-#         print(f"✅ Successfully saved application {app_data.get('applicationId')}")
-#         return True
-#     except Exception as e:
-#         print(f"❌ Failed to save application to Weaviate: {e}")
-#         return False
-
-# def save_team_to_weaviate(team_data: dict) -> bool:
-#     """Saves a new team object to the Weaviate 'Team' collection."""
-#     try:
-#         teams = weaviate_client.collections.get("Team")
-#         teams.data.insert(properties=team_data)
-#         print(f"✅ Successfully saved team {team_data.get('teamId')}")
-#         return True
-#     except Exception as e:
-#         print(f"❌ Failed to save team to Weaviate: {e}")
-#         return False
-
-# def update_team_add_member(team_name: str, member_name: str, skills: list) -> str:
-#     """Finds a team by name, adds a member and their skills, then updates it."""
-#     try:
-#         teams = weaviate_client.collections.get("Team")
-#         # Step 1: Find the team by its name
-#         response = teams.query.fetch_objects(
-#             filters=Filter.by_property("name").equal(team_name),
-#             limit=1
-#         )
-#         if not response.objects:
-#             return f"⚠️ Team '{team_name}' not found."
-
-#         team_obj = response.objects[0]
-        
-#         # Step 2: Update the properties
-#         updated_members = list(set((team_obj.properties.get("members") or []) + [member_name]))
-#         updated_skills = list(set((team_obj.properties.get("skills") or []) + skills))
-        
-#         # Step 3: Replace the object with the updated data
-#         teams.data.replace(
-#             uuid=team_obj.uuid,
-#             properties={
-#                 **team_obj.properties,
-#                 "members": updated_members,
-#                 "skills": updated_skills
-#             }
-#         )
-#         return f"🎉 You have been added to the team '{team_name}'."
-#     except Exception as e:
-#         print(f"❌ Failed to update team in Weaviate: {e}")
-#         return "⚠️ An error occurred while trying to join the team."
-
-# # -------------------- Main message handler (state machine) --------------------
-# def handle_user_message(session: dict, user_text: str, uploaded_file: Any = None) -> (str, dict, bool):
-#     session = session or initial_session()
-#     st = session.get("state", "idle")
-#     text = (user_text or "").strip()
-
-#     if text.lower() in ("cancel", "exit", "quit", "restart", "reset"):
-#         return "Conversation reset. How can I help?", initial_session(), False
-
-#     if uploaded_file:
-#         bot_msg, new_session = handle_uploaded_cv_for_session(session, uploaded_file)
-#         return bot_msg, new_session, False
-
-#     # ========== IDLE STATE ==========
-#     if st == "idle":
-#         low = text.lower()
-#         if any(k in low for k in ["hi", "hello", "hey", "مرحبا"]):
-#             return "👋 Hello! How can I support you today? You can ask about jobs, teams, or recommendations.", session, False
-        
-#         # Flow starters
-#         if any(k in low for k in ["apply", "أريد التقديم"]):
-#             session["state"] = "apply_name"
-#             return "Okay, let's start your application. What's your full name?", initial_session(), False
-#         if any(k in low for k in ["team", "فريق"]):
-#             session["state"] = "team_action"
-#             return "Do you want to 'create' a team or 'join' an existing one?", initial_session(), False
-#         if any(k in low for k in ["recommend", "رشح"]):
-#             session["state"] = "recommend_wait_cv"
-#             return "Please upload your CV to get job recommendations.", initial_session(), True
-
-#         # Intent Routing for RAG/KB
-#         intent = route_intent(text)
-#         if intent:
-#             if intent.startswith("kb_"):
-#                 return kb_fallback(intent), session, False
-#             if intent.startswith("rag_"):
-#                 class_map = {"rag_job": "Job", "rag_project": "Project", "rag_opportunities": "Opportunities"}
-#                 rag_ans, _ = rag_answer(text, class_map[intent])
-#                 return rag_ans, session, False
-
-#         # Fallback to general LLM
-#         try:
-#             resp = llm_client.chat.completions.create(model=MODEL_NAME, messages=[{"role": "system", "content": SYSTEM_PROMPT_BASE}, {"role": "user", "content": text}], temperature=0.3, max_tokens=4096)
-#             return resp.choices[0].message.content or "...", session, False
-#         except Exception as e:
-#             return f"⚠️ Sorry, an error occurred: {e}", session, False
-
-#     # ========== APPLY FLOW ==========
-#     if st == "apply_name":
-#         session["data"]["applicantName"] = text
-#         session["state"] = "apply_email"
-#         return "Thanks. What's your email address?", session, False
-#     if st == "apply_email":
-#         session["data"]["applicantEmail"] = text
-#         session["state"] = "apply_cover"
-#         return "Got it. Please type a short cover letter (or 'skip').", session, False
-#     if st == "apply_cover":
-#         session["data"]["coverLetter"] = "" if text.lower() == "skip" else text
-#         session["state"] = "apply_wait_cv"
-#         return "Please upload your CV now.", session, True
-#     if st == "apply_jobtitle":
-#         session["data"]["targetJobTitle"] = text
-#         session["state"] = "apply_confirm"
-#         return f"You are applying for '{text}'. Is this correct? (yes/no)", session, False
-#     if st == "apply_confirm":
-#         if text.lower() in ("yes", "y", "نعم"):
-#             app_data = {
-#                 "applicationId": str(uuid.uuid4()),
-#                 "jobId": session["data"].get("targetJobTitle"), # Simplified to title for now
-#                 "applicantName": session["data"].get("applicantName"),
-#                 "applicantEmail": session["data"].get("applicantEmail"),
-#                 "coverLetter": session["data"].get("coverLetter", ""),
-#                 "cvText": session["data"].get("cvText", ""),
-#                 "skills": session["data"].get("cvSkills", []),
-#                 "createdAt": get_rfc3339_time()
-#             }
-#             ok = save_application_to_weaviate(app_data)
-#             msg = "🎉 Your application has been submitted successfully!" if ok else "⚠️ Failed to save application."
-#             return msg, initial_session(), False
-#         else:
-#             return "Application cancelled. How else can I help?", initial_session(), False
-
-#     # ========== TEAM FLOW ==========
-#     if st == "team_action":
-#         if "create" in text.lower(): session["state"] = "team_create_name"; return "Great! What's the team name?", session, False
-#         if "join" in text.lower(): session["state"] = "team_join_name"; return "Okay, what's the name of the team to join?", session, False
-#         return "Please say 'create' or 'join'.", session, False
-#     if st == "team_create_name":
-#         session["data"]["team_name"] = text; session["state"] = "team_create_owner"; return "Team name saved. Who is the owner?", session, False
-#     if st == "team_create_owner":
-#         session["data"]["owner"] = text; session["state"] = "team_create_skills"; return "Owner saved. List the team's skills (comma-separated).", session, False
-#     if st == "team_create_skills":
-#         session["data"]["skills"] = [s.strip() for s in text.split(",")]; session["state"] = "team_create_idea"; return "Skills saved. Briefly describe the project idea.", session, False
-#     if st == "team_create_idea":
-#         team_data = {
-#             "teamId": str(uuid.uuid4()), "name": session["data"].get("team_name"),
-#             "members": [session["data"].get("owner")], "skills": session["data"].get("skills", []),
-#             "creatorId": session["data"].get("owner"), "idea": text, "createdAt": get_rfc3339_time()
-#         }
-#         saved = save_team_to_weaviate(team_data)
-#         msg = f"🎉 Team '{team_data['name']}' created!" if saved else "⚠️ Failed to create team."
-#         return msg, initial_session(), False
-#     if st == "team_join_name":
-#         session["data"]["team_name"] = text; session["state"] = "team_join_member"; return "What's your name (to add to the team)?", session, False
-#     if st == "team_join_member":
-#         session["data"]["member_name"] = text; session["state"] = "team_join_skills"; return "Enter your skills (comma-separated).", session, False
-#     if st == "team_join_skills":
-#         skills = [s.strip() for s in text.split(",") if s.strip()]
-#         resp = update_team_add_member(session["data"].get("team_name"), session["data"].get("member_name"), skills)
-#         return resp, initial_session(), False
-        
-#     # Fallback
-#     return "Sorry, I'm not sure how to handle that. Let's start over.", initial_session(), False
-
-
-# # ================================
-# # Part 4 — Gradio Chat UI wiring
-# # ================================
-# atexit.register(lambda: weaviate_client.close())
-
-# with gr.Blocks(css="""
-# .chatbot {height: 520px; overflow: auto;}
-# .user-bubble {background-color: #DCF8C6; padding: 10px; border-radius: 12px; max-width: 75%; float: right; clear: both; margin: 5px; word-wrap: break-word;}
-# .bot-bubble {background-color: #F1F0F0; padding: 10px; border-radius: 12px; max-width: 75%; float: left; clear: both; margin: 5px; word-wrap: break-word;}
-# .chatbox-container {display: flex; gap: 8px; margin-top: 10px;}
-# """) as demo:
-#     gr.Markdown("# 💬 EduNatives — Conversational Job Portal")
-#     chat_html = gr.HTML(format_chat_html([]))
-#     with gr.Row(elem_classes="chatbox-container"):
-#         user_input = gr.Textbox(placeholder="Type your message (e.g., 'apply', 'create team', 'recommend')", scale=4)
-#         send_btn = gr.Button("Send", variant="primary", scale=1)
-#     with gr.Row(visible=False) as file_row:
-#         cv_uploader = gr.File(label="Upload CV (.pdf/.docx/.txt)", file_count="single", scale=4)
-#         upload_btn = gr.Button("Upload CV", scale=1, visible=True) # Button is visible within the row
-#     with gr.Row():
-#         clear_btn = gr.Button("Reset Conversation")
-    
-#     chat_history_state = gr.State([])        
-#     session_state = gr.State(initial_session)
-
-#     def handle_send(message: str, history: List[Dict[str, str]], session: dict):
-#         history.append({"role": "user", "content": message})
-#         bot_reply, new_session, show_uploader = handle_user_message(session, message, None)
-#         history.append({"role": "assistant", "content": bot_reply})
-#         return "", format_chat_html(history), history, new_session, gr.update(visible=show_uploader), gr.update(visible=show_uploader)
-
-#     def handle_upload(file_obj, history: List[Dict[str, str]], session: dict):
-#         if not file_obj:
-#             history.append({"role": "assistant", "content": "Please select a file first."})
-#             return format_chat_html(history), history, session, gr.update(visible=True), gr.update(visible=True)
-            
-#         history.append({"role": "user", "content": f"📎 Uploaded: {os.path.basename(file_obj.name)}"})
-#         bot_reply, new_session, show_uploader = handle_user_message(session, "", file_obj)
-#         history.append({"role": "assistant", "content": bot_reply})
-#         return format_chat_html(history), history, new_session, gr.update(visible=show_uploader), gr.update(visible=show_uploader)
-
-#     def handle_reset():
-#         return format_chat_html([]), [], initial_session(), gr.update(visible=False), gr.update(visible=False)
-
-#     send_btn.click(
-#         fn=handle_send,
-#         inputs=[user_input, chat_history_state, session_state],
-#         outputs=[user_input, chat_html, chat_history_state, session_state, file_row, cv_uploader]
-#     )
-#     user_input.submit(
-#         fn=handle_send,
-#         inputs=[user_input, chat_history_state, session_state],
-#         outputs=[user_input, chat_html, chat_history_state, session_state, file_row, cv_uploader]
-#     )
-#     upload_btn.click(
-#         fn=handle_upload,
-#         inputs=[cv_uploader, chat_history_state, session_state],
-#         outputs=[chat_html, chat_history_state, session_state, file_row, cv_uploader]
-#     )
-#     clear_btn.click(
-#         fn=handle_reset,
-#         inputs=[],
-#         outputs=[chat_html, chat_history_state, session_state, file_row, cv_uploader]
-#     )
-
-# if __name__ == "__main__":
-#     demo.launch(debug=True)
-
-
-
-
-from __future__ import annotations
 import os
-import re
-import uuid
+import weaviate
+from weaviate.auth import Auth
+from openai import OpenAI
 import json
-import time
-import fitz  # PyMuPDF
-import docx
-import markdown
-from datetime import datetime, timezone
-from typing import List, Dict, Any, Optional
-from dataclasses import dataclass
-
 import gradio as gr
-import numpy as np
-from openai import OpenAI
-import weaviate
-from weaviate.classes.init import Auth
-from weaviate.classes.config import Configure, Property, DataType
-from weaviate.classes.query import Filter
-
-# -------------------- Configuration --------------------
-MODEL_NAME = os.getenv("MODEL_NAME", "openai/gpt-oss-120b")
-DEEPINFRA_API_KEY = os.getenv("DEEPINFRA_API_KEY", "285LUJulGIprqT6hcPhiXtcrphU04FG4")
-BASE_URL = os.getenv("BASE_URL", "https://api.deepinfra.com/v1/openai")
-
-WEAVIATE_URL = os.getenv("WEAVIATE_URL", "htorgbgpt4w63nvf1yeuw.c0.us-west3.gcp.weaviate.cloud")
-WEAVIATE_API_KEY = os.getenv("WEAVIATE_API_KEY", "ZUd6clB5WmYzVGkxeU40cl96NTY5UkViUlVzY05Md3IzQ0JKelBZQmxGZHRPeGpCeGdxS1FUNnlYUkFFPV92MjAw")
-
-MEMORY_FILE = os.getenv("MEMORY_FILE", "chat_memory.json")
-LOG_FILE = os.getenv("LOG_FILE", "interaction_logs.json")
-
-# -------------------- Clients --------------------
-llm_client = OpenAI(api_key=DEEPINFRA_API_KEY, base_url=BASE_URL)
-
-weaviate_client = weaviate.connect_to_weaviate_cloud(
-    cluster_url=WEAVIATE_URL,
-    auth_credentials=Auth.api_key(WEAVIATE_API_KEY),
-    skip_init_checks=True
-)
-
-# -------------------- Helpers --------------------
-UTC = timezone.utc
-
-SYSTEM_PROMPT_BASE = (
-    "You are *EduNatives Assistant*, a helpful, friendly, and precise academic/career guide. "
-    "You serve three primary audiences: Students, Universities/Researchers, and Companies.\n\n"
-    "Goals by audience:\n"
-    "- Students: registration/account help; finding internships/scholarships; connecting with mentors or professors; querying projects; applying for jobs/opportunities; joining project teams.\n"
-    "- Universities/Researchers: publish research or announce events; connect/collaborate with students.\n"
-    "- Companies: post jobs/internships/graduate roles; discover student talent.\n\n"
-    "General rules:\n"
-    "- Reply in the user's language (Arabic if the user writes Arabic; otherwise English).\n"
-    "- Use data from Weaviate collections (Job, Opportunities, Project) when relevant.\n"
-    "- Be concise, step-by-step, and action-oriented (lists, bullets, checklists).\n"
-    "- If information is unavailable, state that clearly and suggest the next best step.\n"
-    "- For CV analysis, extract skills/experience and recommend matching opportunities.\n"
-    "- Assist with applications and team matching.\n"
-    "- Ensure that all generated prompts are phrased using positive reinforcement."
-)
-
-def get_rfc3339_time() -> str:
-    """Return current UTC time in RFC3339 format"""
-    return datetime.now(UTC).isoformat().replace('+00:00', 'Z')
-
-# --- CV Skills Extraction (Regex baseline, can replace with NLP model later) ---
-_SKILL_REGEX = re.compile(r"\b(Natural Language Processing|Building Information Modeling|Search Engine Optimization|Search Engine Marketing|Aerospace Engineering & Management|Computational Fluid Dynamics|Kotlin Multiplatform|Google Cloud Platform|Social Media Marketing|Aerospace Engineering|Microsoft SQL Server|Amazon Web Services|Finite Element Analysis|Technology-based Management|Autodesk Inventor|Emotional Intelligence|Aerospace Engineering & Operations Management|Content Marketing|Presentation Skills|Interpersonal Skills|Critical Thinking|Financial Modeling|Decision Making|Process Improvement|Time Management|Lean Manufacturing|Project Management|Microsoft Excel|Data Visualization|Computer Vision|Machine Learning|Deep Learning|Attention to Detail|Six Sigma|Risk Analysis|Data Analysis|Data Science|Communication|Collaboration|Teamwork|Leadership|Management|Adaptability|Creativity|Innovation|Negotiation|Android|Angular|Ansible|Apache|ArcGIS|Arduino|Asana|ASP\.NET|AutoCAD|Azure|Bash|BIM|Business Analysis|C\+\+|C#|CAM|Cassandra|CATIA|CentOS|Chef|CI/CD|Civil 3D|Content Marketing|CRM|CSS|Data Mining|Django|Docker|Elasticsearch|Email Marketing|ERP|ETABS|ETL|Express\.js|Facebook Ads|Firebase|Flask|Flutter|FPGA|Fusion 360|GCP|Git|GitHub|GitLab|Go|Google Ads|Google Analytics|GraphQL|Hadoop|HTML|HubSpot|iOS|Java|JavaScript|Jenkins|Jira|Jupyter Notebook|Kanban|Keras|Kotlin|Kubernetes|LabVIEW|Laravel|LESS|Linux|Lua|macOS|Marketo|MATLAB|Matplotlib|MongoDB|Multisim|MySQL|Nginx|NLP|Node\.js|NoSQL|Nuxt\.js|NumPy|Next\.js|Objective-C|Oracle Database|Oracle|OrCAD|Pandas|Perl|PHP|PLC|Plotly|PostgreSQL|Power BI|PowerShell|Problem Solving|Puppet|PSpice|Python|PyTorch|Raspberry Pi|React Native|React|Red Hat|Redis|Revit|Ruby on Rails|Ruby|Rust|Salesforce|SAP2000|SAP|Sass|SCADA|Scala|Scikit-learn|Scrum|Seaborn|SEM|SEO|Simulink|SketchUp|Slack|SolidWorks|Spring Boot|SQL|SQLAlchemy|SwiftUI|Swift|Tableau|Terraform|TensorFlow|Trello|TypeScript|Ubuntu|Verilog|VHDL|Vue\.js|Waterfall|Windows|WordPress|Xamarin|Analytical Skills)\b", re.IGNORECASE)
-
-def extract_skills_from_text(cv_text: str) -> List[str]:
-    skills = list({m.group(0).lower() for m in _SKILL_REGEX.finditer(cv_text)})
-    return [s.capitalize() for s in skills]
-
-# --- Process uploaded file (PDF, DOCX, TXT) ---
-def process_uploaded_file(file_obj: Any) -> dict | None:
-    if not file_obj:
-        return None
-    file_path = file_obj.name
-    filename = os.path.basename(file_path)
-    text_content = ""
-    try:
-        if filename.lower().endswith(".pdf"):
-            with fitz.open(file_path) as doc:
-                for page in doc:
-                    text_content += page.get_text()
-        elif filename.lower().endswith(".docx"):
-            docp = docx.Document(file_path)
-            for p in docp.paragraphs:
-                text_content += p.text + "\n"
-        elif filename.lower().endswith(".txt"):
-            with open(file_path, "r", encoding="utf-8") as f:
-                text_content = f.read()
-        else:
-            return {"error": f"Unsupported file type: {filename}"}
-
-        skills = extract_skills_from_text(text_content)
-        return {"content": text_content.strip(), "skills": skills, "filename": filename}
-
-    except Exception as e:
-        return {"error": f"Error processing file {filename}: {e}"}
-
-ARABIC_RANGE = (
-    (0x0600, 0x06FF), (0x0750, 0x077F), (0x08A0, 0x08FF),
-    (0xFB50, 0xFDFF), (0xFE70, 0xFEFF), (0x1EE00, 0x1EEFF)
-)
-
-def is_arabic(text: str) -> bool:
-    for ch in text:
-        code = ord(ch)
-        for a, b in ARABIC_RANGE:
-            if a <= code <= b:
-                return True
-    return False
-
-# --- Chat history HTML formatter (for Gradio) ---
-def format_chat_html(history: List[Dict[str, str]]) -> str:
-    html = "<div class='chatbot'>"
-    for msg in history:
-        role = msg["role"]
-        content = msg["content"]
-        if role == "user":
-            html += f"<div class='user-bubble'>{content}</div>"
-        else:
-            html_content = markdown.markdown(content, extensions=['tables'])
-            html += f"<div class='bot-bubble'>{html_content}</div>"
-    html += "</div>"
-    return html
-
-# ================================
-# Part 2 — Weaviate Collections + Query + RAG
-# ================================
-
-# -------------------- Clean Query Text --------------------
-def clean_query_text(query_text: str) -> str:
-    """Clean the query text to extract key terms for better search results."""
-    stop_words = ["show", "me", "available", "jobs", "job", "please", "find", "search", "for"]
-    query_text = query_text.lower().strip()
-    for word in stop_words:
-        query_text = query_text.replace(f" {word} ", " ")
-    query_text = re.sub(r'[^\w\s]', '', query_text)
-    query_text = " ".join(query_text.split())
-    return query_text
-
-# -------------------- Ensure collections --------------------
-def ensure_collections():
-    if not weaviate_client.collections.exists("Application"):
-        weaviate_client.collections.create(
-            name="Application",
-            properties=[
-                Property(name="applicationId", data_type=DataType.TEXT),
-                Property(name="jobId", data_type=DataType.TEXT),
-                Property(name="applicantName", data_type=DataType.TEXT),
-                Property(name="applicantEmail", data_type=DataType.TEXT),
-                Property(name="coverLetter", data_type=DataType.TEXT),
-                Property(name="cvText", data_type=DataType.TEXT),
-                Property(name="skills", data_type=DataType.TEXT_ARRAY),
-                Property(name="createdAt", data_type=DataType.DATE),
-            ],
-            vectorizer_config=Configure.Vectorizer.none()
-        )
+import atexit
+import datetime
+import re 
 
-    if not weaviate_client.collections.exists("Team"):
-        weaviate_client.collections.create(
-            name="Team",
-            properties=[
-                Property(name="teamId", data_type=DataType.TEXT),
-                Property(name="name", data_type=DataType.TEXT),
-                Property(name="projectId", data_type=DataType.TEXT),
-                Property(name="members", data_type=DataType.TEXT_ARRAY),
-                Property(name="skills", data_type=DataType.TEXT_ARRAY),
-                Property(name="createdAt", data_type=DataType.DATE),
-                Property(name="creatorId", data_type=DataType.TEXT),
-            ],
-            vectorizer_config=Configure.Vectorizer.none()
-        )
+import pypdf
+import docx
 
-    if not weaviate_client.collections.exists("Memory"):
-        weaviate_client.collections.create(
-            name="Memory",
+MODEL_NAME = "openai/gpt-oss-120b"
+EMBEDDING_MODEL_NAME = "Qwen/Qwen3-Embedding-8B"
+DEEPINFRA_API_KEY = "KwZiFcFHhOPUE6Rrc6wY4ng0mqPfwsVN"
+BASE_URL = "https://api.deepinfra.com/v1/openai"
+WEAVIATE_URL = "https://maf5cvz1saelnti3k34a.c.europe-west3.gcp.weaviate.cloud"
+WEAVIATE_API_KEY = "L1U0ZndLYXFuSWVsL3NIT19NTzArZFZNU3BsaUlabUlEOHNMaDIrZVRwU2t2OTJTekFxRDZTYTBlTjc0PV92MjAw"
+
+def create_application_schema(client: weaviate.WeaviateClient):
+    collection_name = "Application"
+    if not client.collections.exists(collection_name):
+        print(f"Creating collection: {collection_name}")
+        client.collections.create(
+            name=collection_name,
             properties=[
-                Property(name="memoryId", data_type=DataType.TEXT),
-                Property(name="sessionId", data_type=DataType.TEXT),
-                Property(name="text", data_type=DataType.TEXT),
-                Property(name="createdAt", data_type=DataType.DATE),
-            ],
-            vectorizer_config=Configure.Vectorizer.none()
+                weaviate.classes.config.Property(name="job_id", data_type=weaviate.classes.config.DataType.TEXT),
+                weaviate.classes.config.Property(name="cv_content", data_type=weaviate.classes.config.DataType.TEXT),
+                weaviate.classes.config.Property(name="cover_letter_content", data_type=weaviate.classes.config.DataType.TEXT),
+                weaviate.classes.config.Property(name="submission_date", data_type=weaviate.classes.config.DataType.DATE),
+            ]
         )
-
-    if not weaviate_client.collections.exists("Job"):
-        weaviate_client.collections.create(
-            name="Job",
-            properties=[
-                Property(name="jobId", data_type=DataType.TEXT),
-                Property(name="title", data_type=DataType.TEXT),
-                Property(name="companyName", data_type=DataType.TEXT),
-                Property(name="description", data_type=DataType.TEXT),
-                Property(name="skills", data_type=DataType.TEXT_ARRAY),
-                Property(name="workplaceType", data_type=DataType.TEXT),
-                Property(name="employmentType", data_type=DataType.TEXT_ARRAY),
-                Property(name="requirements", data_type=DataType.TEXT),
-            ],
-            vectorizer_config=Configure.Vectorizer.text2vec_openai()
-        )
-
-    if not weaviate_client.collections.exists("Opportunities"):
-        weaviate_client.collections.create(
-            name="Opportunities",
-            properties=[
-                Property(name="opportunityId", data_type=DataType.TEXT),
-                Property(name="title", data_type=DataType.TEXT),
-                Property(name="organization", data_type=DataType.TEXT),
-                Property(name="description", data_type=DataType.TEXT),
-                Property(name="type", data_type=DataType.TEXT),
-                Property(name="skills", data_type=DataType.TEXT_ARRAY),
-            ],
-            vectorizer_config=Configure.Vectorizer.text2vec_openai()
+        print(f"✅ Collection '{collection_name}' created successfully.")
+    else:
+        print(f"✅ Collection '{collection_name}' already exists.")
+
+class WeaviateChatbot:
+    def __init__(self, weaviate_url, weaviate_api_key, llm_api_key, llm_base_url):
+        print("Connecting to clients...")
+        self.weaviate_client = weaviate.connect_to_weaviate_cloud(
+            cluster_url=weaviate_url,
+            auth_credentials=Auth.api_key(weaviate_api_key),
+            skip_init_checks=True
         )
+        self.weaviate_client.connect()
+        print("✅ Successfully connected to Weaviate.")
+        
+        create_application_schema(self.weaviate_client)
 
-    if not weaviate_client.collections.exists("Project"):
-        weaviate_client.collections.create(
-            name="Project",
-            properties=[
-                Property(name="projectId", data_type=DataType.TEXT),
-                Property(name="title", data_type=DataType.TEXT),
-                Property(name="description", data_type=DataType.TEXT),
-                Property(name="skills", data_type=DataType.TEXT_ARRAY),
-                Property(name="creatorName", data_type=DataType.TEXT),
-            ],
-            vectorizer_config=Configure.Vectorizer.text2vec_openai()
-        )
+        self.llm_client = OpenAI(api_key=llm_api_key, base_url=llm_base_url)
+        print("✅ Successfully connected to LLM client (DeepInfra).")
 
-ensure_collections()
+        self.collection_names = ["Job", "Opportunities", "Project"]
 
-# -------------------- Query Weaviate (Hybrid Search) --------------------
-def query_weaviate_collection(class_name: str, query_text: str, limit: int = 10) -> List[dict]:
-    """
-    Performs a hybrid search on a Weaviate collection to get more relevant results
-    for conversational queries.
-    """
-    try:
-        print(f"[DEBUG] Querying collection {class_name} with query: {query_text}")
-        collection = weaviate_client.collections.get(class_name)
-        cleaned_query = clean_query_text(query_text)
-        print(f"[DEBUG] Cleaned query: {cleaned_query}")
-        response = collection.query.hybrid(
-            query=cleaned_query,
-            limit=limit,
-            query_properties=["title^2", "description", "skills"] if class_name in ["Job", "Project", "Opportunities"] else None
-        )
-        items = [obj.properties for obj in response.objects]
-        print(f"[DEBUG] Found {len(items)} results for {class_name}: {items}")
-        return items
-    except Exception as e:
-        print(f"[Weaviate Query Error] {class_name} - {e}")
-        return []
+    def _embed_text(self, text: str) -> list[float]:
+        resp = self.llm_client.embeddings.create(model=EMBEDDING_MODEL_NAME, input=text, encoding_format="float")
+        return resp.data[0].embedding
 
-# -------------------- Search All Collections --------------------
-def search_all_collections(query_text: str, limit_per_class: int = 3) -> list[dict]:
-    """
-    Searches across Job, Project, and Opportunities collections and returns
-    a combined list of results, with each result tagged by its class name.
-    """
-    collections_to_search = ["Job", "Project", "Opportunities"]
-    all_results = []
-    for class_name in collections_to_search:
-        items = query_weaviate_collection(class_name, query_text, limit=limit_per_class)
-        for item_properties in items:
-            all_results.append({
-                "class_name": class_name,
-                "properties": item_properties
+    def _search_database(self, query_vector: list[float], limit: int = 5) -> str:
+        all_results = []
+        for name in self.collection_names:
+            try:
+                collection = self.weaviate_client.collections.get(name)
+                response = collection.query.near_vector(near_vector=query_vector, limit=limit)
+                for item in response.objects:
+                    all_results.append(f"Type: {name}\nContent: {json.dumps(item.properties, indent=2, default=str)}\n")
+            except Exception as e:
+                print(f"Could not query collection '{name}'. Error: {e}")
+        return "\n---\n".join(all_results) if all_results else "No relevant information found in the database."
+    
+    def _generate_response(self, query: str, context: str) -> str:
+        prompt = f"""
+You are *EduNatives Assistant*. Your main goal is to help users find opportunities and apply for them.
+- First, answer the user's question based on the CONTEXT from the database.
+- **IMPORTANT**: If you list jobs, make sure each job has a clear identifier like (job_021).
+- If the user says they want to apply for a job, for example "I want to apply for job_021", you MUST respond ONLY with the exact phrase: `STARTING_APPLICATION_PROCESS:job_021`. Do not add any other text.
+
+--- CONTEXT FROM DATABASE START ---
+{context}
+--- CONTEXT FROM DATABASE END ---
+
+User Question: {query}
+
+Answer:
+"""
+        response = self.llm_client.chat.completions.create(model=MODEL_NAME, messages=[{"role": "user", "content": prompt}], max_tokens=4096)
+        return response.choices[0].message.content.strip()
+
+    def ask(self, query: str):
+        print(f"\nProcessing query: '{query}'")
+        query_vector = self._embed_text(query)
+        context = self._search_database(query_vector)
+        answer = self._generate_response(query, context)
+        return answer
+
+    def save_application(self, application_data: dict):
+        print("Saving application to Weaviate...")
+        try:
+            applications = self.weaviate_client.collections.get("Application")
+            app_uuid = applications.data.insert({
+                "job_id": application_data.get("job_id"),
+                "cv_content": application_data.get("cv_content"),
+                "cover_letter_content": application_data.get("cover_letter_content"),
+                "submission_date": datetime.datetime.now(datetime.timezone.utc)
             })
-    return all_results
-
-# -------------------- RAG Answer (Multi-Class Search) --------------------
-def rag_answer_all(user_question: str, top_k: int = 3) -> (str, list[dict]):
-    retrieved_items = search_all_collections(user_question, limit_per_class=top_k)
+            print(f"✅ Application saved with UUID: {app_uuid}")
+            return True
+        except Exception as e:
+            print(f"❌ Failed to save application: {e}")
+            return False
+
+    def close_connections(self):
+        if self.weaviate_client.is_connected():
+            self.weaviate_client.close()
+            print("\nWeaviate connection closed.")
+
+
+# --- Helper to extract text from uploaded files ---
+def _extract_text_from_file(file_path):
+    # ... (No changes needed in this function, code is omitted for brevity)
+    print(f"Extracting text from: {file_path}")
+    if file_path.endswith('.pdf'):
+        try:
+            reader = pypdf.PdfReader(file_path)
+            text = "".join(page.extract_text() for page in reader.pages)
+            return text
+        except Exception as e:
+            return f"Error reading PDF: {e}"
+    elif file_path.endswith('.docx'):
+        try:
+            doc = docx.Document(file_path)
+            return "\n".join([para.text for para in doc.paragraphs])
+        except Exception as e:
+            return f"Error reading DOCX: {e}"
+    elif file_path.endswith('.txt'):
+        try:
+            with open(file_path, 'r', encoding='utf-8') as f:
+                return f.read()
+        except Exception as e:
+            return f"Error reading TXT: {e}"
+    return "Unsupported file type."
+
+
+chatbot_instance = WeaviateChatbot(WEAVIATE_URL, WEAVIATE_API_KEY, DEEPINFRA_API_KEY, BASE_URL)
+atexit.register(chatbot_instance.close_connections)
+
+def chat_interface_func(message: str, history: list, app_state: dict, file_obj: object):
+    history = history or []
     
-    if not retrieved_items:
-        return f"Sorry, I couldn't find any results related to '{user_question}' in our Jobs, Projects, or Opportunities databases.", []
-
-    context_parts = []
-    grouped_results = {}
-    for item in retrieved_items:
-        class_name = item["class_name"]
-        if class_name not in grouped_results:
-            grouped_results[class_name] = []
-        grouped_results[class_name].append(item["properties"])
-
-    for class_name, items in grouped_results.items():
-        context_parts.append(f"\n--- Results from '{class_name}' collection ---")
-        for i, properties in enumerate(items, 1):
-            details = {k: str(v) for k, v in properties.items()}
-            item_str = f"Record {i}:\n{json.dumps(details, indent=2, ensure_ascii=False)}"
-            context_parts.append(item_str)
+    if file_obj is not None:
+        file_path = file_obj.name
+        text = _extract_text_from_file(file_path)
+        
+        if app_state.get("mode") == "APPLYING_CV":
+            app_state["cv_content"] = text
+            history.append((f"📄 CV '{os.path.basename(file_path)}' uploaded.", "Great! Now, please upload your Cover Letter."))
+            app_state["mode"] = "APPLYING_COVER_LETTER"
+            return history, app_state, gr.update(visible=True, value=None)
+        
+        elif app_state.get("mode") == "APPLYING_COVER_LETTER":
+            app_state["cover_letter_content"] = text
+            history.append((f"📄 Cover Letter '{os.path.basename(file_path)}' uploaded.", "Thank you! Submitting your application now..."))
             
-    context_block = "\n".join(context_parts)
-    prompt = f"""
-    User Question: "{user_question}"
-    You are an expert AI assistant. Your mission is to analyze structured data from different categories (Jobs, Projects, Opportunities) and present a comprehensive, clear summary to the user.
-    *Primary Directive:* Your ONLY source of information is the structured JSON data provided below under "Retrieved Data". If the data section is empty, state that no results were found.
-    *Your Core Instructions:*
-    1.  *Acknowledge the Categories:* Analyze all the data provided from each collection (Job, Project, Opportunities).
-    2.  *Summarize Logically:* For each result, *you must clearly state which category it belongs to. For example, start with "I found a **Job* opportunity:" or "Here is a *Project* you might be interested in:".
-    3.  *Present All Details:* Convert the data for each item into natural, readable language, covering all important details like title, company/creator, description, and skills.
-    4.  *Use Clear Formatting:* Use Markdown headings (e.g., ### Job: [Title]) and bullet points to make the response easy to read.
-    Retrieved Data:
-    {context_block}
-    """
-
-    try:
-        resp = llm_client.chat.completions.create(
-            model=MODEL_NAME,
-            messages=[
-                {"role": "system", "content": SYSTEM_PROMPT_BASE},
-                {"role": "user", "content": prompt}
-            ],
-            temperature=0.3,
-            max_tokens=4096
-        )
-        answer = resp.choices[0].message.content or ""
-    except Exception as e:
-        print(f"[RAG LLM Error] {e}")
-        answer = "⚠️ Sorry, I couldn't process that. Try again later."
-
-    return answer, retrieved_items
-
-# ================================
-# Part 3 — Conversation State Machine + Embedding Recommendations
-# ================================
-
-# -------------------- Embedding helpers --------------------
-def compute_embedding(text: str) -> List[float]:
-    try:
-        resp = llm_client.embeddings.create(
-            model="Qwen/Qwen3-Embedding-8B",
-            input=text,
-            encoding_format="float"
-        )
-        if hasattr(resp, "data") and resp.data:
-            return resp.data[0].embedding
-    except Exception as e:
-        print("[compute_embedding] error:", e)
-    return []
-
-def cosine_similarity(a: List[float], b: List[float]) -> float:
-    try:
-        va = np.array(a, dtype=float)
-        vb = np.array(b, dtype=float)
-        if va.size == 0 or vb.size == 0: return 0.0
-        denom = (np.linalg.norm(va) * np.linalg.norm(vb))
-        if denom == 0: return 0.0
-        return float(np.dot(va, vb) / denom)
-    except Exception as e:
-        print("[cosine_similarity] error:", e)
-        return 0.0
-
-# -------------------- Recommendations by embedding --------------------
-def recommend_jobs_by_embedding(cv_text: str, top_k: int = 5, jobs_fetch_limit: int = 200) -> str:
-    skills = extract_skills_from_text(cv_text or "")
-    user_text = " ".join(skills) if skills else (cv_text or "")[:500]
-    user_emb = compute_embedding(user_text)
-    if not user_emb:
-        return "⚠️ Unable to compute embedding for your CV. Try again or check API keys."
-
-    try:
-        jobs_col = weaviate_client.collections.get("Job")
-        fetched = jobs_col.query.fetch_objects(limit=jobs_fetch_limit)
-        if not fetched.objects:
-            return "⚠️ No jobs found in the database."
-    except Exception as e:
-        print("[recommend_jobs_by_embedding] Weaviate fetch error:", e)
-        return "⚠️ Could not fetch jobs from the database."
-
-    scored_jobs = []
-    for obj in fetched.objects:
-        props = obj.properties
-        job_text_parts = []
-        if props.get("skills"): job_text_parts.append(" ".join(props.get("skills")))
-        if props.get("title"): job_text_parts.append(props.get("title"))
-        if props.get("description"): job_text_parts.append((props.get("description") or "")[:2000])
-        job_text = " ".join(job_text_parts).strip() or (props.get("title") or "")
-        job_emb = compute_embedding(job_text)
-        if not job_emb: continue
-        score = cosine_similarity(user_emb, job_emb)
-        scored_jobs.append((score, props))
-
-    if not scored_jobs:
-        return "⚠️ No jobs could be embedded / compared."
-    
-    scored_jobs.sort(key=lambda x: x[0], reverse=True)
-    top = scored_jobs[:top_k]
-    lines = []
-    for score, props in top:
-        title = props.get("title", "No title")
-        company = props.get("companyName", "Unknown company")
-        job_id = props.get("jobId", "")
-        salary = props.get("salary") or props.get("salaryDetails") or "Not specified"
-        skills_list = props.get("skills") or []
-        description = (props.get("description") or "").strip()
-        lines.append(
-            f"*{title}* at {company} \n"
-            f"- Job ID: {job_id}  \n"
-            f"- Score: {score:.3f}  \n"
-            f"- Salary: {salary}  \n"
-            f"- Skills: {skills_list}  \n"
-            f"- Description: {description[:600]}{'...' if len(description)>600 else ''}  \n"
-            f"---"
-        )
-    return "\n\n".join(lines)
-
-# -------------------- INTENT ROUTING & KNOWLEDGE BASE --------------------
-def route_intent(text: str) -> Optional[str]:
-    low = text.lower()
-    if "تسجيل" in low or "register" in low: return "kb_registration"
-    if ("مين" in low and "edunatives" in low) or ("who is" in low and "edunatives" in low): return "kb_about"
-    if ("كيف" in low and "وظيفة" in low) or ("how to apply" in low and "job" in low): return "kb_apply_job"
-    if "اتواصل" in low and ("دعم" in low or "contact support" in low): return "kb_contact_support"
-    return None
-
-def kb_fallback(intent: str) -> Optional[str]:
-    kb = {
-        "kb_registration": "To register as a student, you need to create an account on our platform. You can do this by visiting our website, clicking on 'Register', and filling out the required information. [Arabic: للتسجيل كطالب، تحتاج إلى إنشاء حساب على منصتنا. يمكنك فعل ذلك بزيارة موقعنا، والنقر على 'تسجيل'، ثم ملء البيانات المطلوبة.]",
-        "kb_about": "EduNatives is an academic and career platform that connects students with opportunities, projects, and mentoring from universities and companies. [Arabic: EduNatives هي منصة أكاديمية ومهنية تربط الطلاب بالفرص والمشاريع والإرشاد من الجامعات والشركات.]",
-        "kb_apply_job": "You can apply for a job directly through the platform. Simply find the job you're interested in, click on 'Apply', and follow the step-by-step process, which includes uploading your CV and writing a cover letter. [Arabic: يمكنك التقديم على وظيفة مباشرة من خلال المنصة. ابحث عن الوظيفة التي تهمك، اضغط على 'قدم الآن'، واتبع الخطوات التي تتضمن تحميل سيرتك الذاتية وكتابة خطاب تقديمي.]",
-        "kb_contact_support": "You can contact our support team via email at support@edunatives.com or by phone at +966 123 4567. [Arabic: يمكنك التواصل مع فريق الدعم عبر البريد الإلكتروني support@edunatives.com أو عبر الهاتف على الرقم +966 123 4567.]"
-    }
-    return kb.get(intent, None)
-
-# -------------------- Conversation Session helpers --------------------
-def initial_session() -> dict:
-    return {"state": "idle", "data": {}}
-
-def handle_uploaded_cv_for_session(session: dict, uploaded_file: Any) -> (str, dict):
-    if not uploaded_file: return "⚠️ No file received.", session
-    doc_info = process_uploaded_file(uploaded_file)
-    if not doc_info or "error" in (doc_info or {}):
-        return f"⚠️ Error processing uploaded CV: {doc_info.get('error') if doc_info else 'unknown error'}", session
-
-    session["data"]["cvText"] = doc_info.get("content", "")
-    session["data"]["cvSkills"] = doc_info.get("skills", [])
-    st = session.get("state")
-    if st == "apply_wait_cv":
-        session["state"] = "apply_jobtitle"
-        detected = session["data"]["cvSkills"]
-        return f"CV received. Detected skills: {detected}. Which job title do you want to apply for? (type job title or 'any')", session
-    if st == "recommend_wait_cv":
-        rec_text = recommend_jobs_by_embedding(session["data"]["cvText"], top_k=5)
-        session = initial_session()
-        return f"Here are recommended jobs based on your CV:\n\n{rec_text}", session
-    return "CV uploaded and processed. What would you like to do next?", session
-
-# -------------------- Main message handler (state machine) --------------------
-def handle_user_message(session: dict, user_text: str, uploaded_file: Any = None) -> (str, dict, bool):
-    session = session or initial_session()
-    st = session.get("state", "idle")
-    text = (user_text or "").strip()
-
-    if text.lower() in ("cancel", "exit", "quit", "restart", "reset"):
-        return "Conversation reset. How can I help you now?", initial_session(), False
-
-    if uploaded_file:
-        bot_msg, new_session = handle_uploaded_cv_for_session(session, uploaded_file)
-        return bot_msg, new_session, False
-
-    # ========== IDLE STATE ==========
-    if st == "idle":
-        low = text.lower()
-
-        if low in ("hi", "hello", "hey", "مرحبا", "ازيك", "السلام عليكم"):
-            return "👋 Hello! How can I support you today? You can ask about jobs, teams, or recommendations.", session, False
-        if low in ["who are you?", "who are you", "انت مين", "من انت"]:
-            return ("👋 I am EduNatives Assistant — your friendly academic and career guide.", session, False)
-
-        if any(k in low for k in ["apply", "أريد التقديم", "عايز اقدم"]):
-            session["state"] = "apply_name"; session["data"] = {}
-            return "Okay — let's start your application. What's your full name?", session, False
-        if any(k in low for k in ["team", "create team", "join team", "انضم", "انشاء فريق"]):
-            session["state"] = "team_action"; session["data"] = {}
-            return "Do you want to create a team or join an existing team? (reply 'create' or 'join')", session, False
-        if any(k in low for k in ["recommend", "recommendation", "jobs for me", "رشح"]):
-            session["state"] = "recommend_wait_cv"; session["data"] = {}
-            return "Please upload your CV to get job recommendations (use the Upload button).", session, True
+            success = chatbot_instance.save_application(app_state)
+            if success:
+                final_message = f"✅ Your application for job **{app_state.get('job_id')}** has been submitted successfully! What else can I help you with?"
+            else:
+                final_message = "❌ Sorry, there was an error submitting your application. Please try again later."
+            
+            history.append((None, final_message))
+            app_state = {"mode": "GENERAL"}
+            return history, app_state, gr.update(visible=False, value=None)
 
-        intent = route_intent(text)
-        if intent and intent.startswith("kb_"):
-            kb_ans = kb_fallback(intent)
-            if kb_ans: return kb_ans, session, False
+    if message:
+        history.append((message, None))
         
-        # Default action: perform a global RAG search
-        rag_ans, _ = rag_answer_all(text)
-        return rag_ans, session, False
+        if "APPLYING" in app_state.get("mode", "GENERAL"):
+            history.append((None, "Please upload the requested document to continue."))
+            return history, app_state, gr.update(visible=True)
 
-    # ========== APPLY FLOW ==========
-    if st == "apply_name":
-        session["data"]["applicantName"] = text or "Applicant"; session["state"] = "apply_email"
-        return "Thanks. What's your email address?", session, False
-    if st == "apply_email":
-        m = re.search(r"([a-zA-Z0-9_.+-]+@[a-zA-Z0-9-]+\.[a-zA-Z0-9-.]+)", text)
-        session["data"]["applicantEmail"] = m.group(1) if m else text; session["state"] = "apply_cover"
-        return "Got it. Please type a short cover letter (or type 'skip' to skip).", session, False
-    if st == "apply_cover":
-        session["data"]["coverLetter"] = "" if text.lower() == "skip" else text; session["state"] = "apply_wait_cv"
-        return "Please upload your CV now (use the Upload button).", session, True
-    if st == "apply_jobtitle":
-        session["data"]["targetJobTitle"] = text; found = query_weaviate_collection("Job", text, limit=3)
-        cv_skills = [s.lower() for s in session["data"].get("cvSkills", [])]
-        if found:
-            job = found[0]; job_skills = [s.lower() for s in (job.get("skills") or [])]
-            overlap = len([s for s in cv_skills if s in job_skills])
-            session["data"]["targetJobId"] = job.get("jobId"); session["state"] = "apply_confirm"
-            if overlap > 0:
-                return (f"I found a job: {job.get('title')} at {job.get('companyName')}. "
-                        f"Detected {overlap} overlapping skills. Do you want to confirm application? (yes/no)"), session, False
-            else:
-                return (f"I found {job.get('title')} at {job.get('companyName')}, but no skill overlap. "
-                        "Proceed anyway? (yes/no)"), session, False
-        else:
-            session["data"]["targetJobId"] = None; session["state"] = "apply_confirm"
-            return f"I couldn't find a job with that title. Apply for '{text}' anyway? (yes/no)", session, False
-    if st == "apply_confirm":
-        if text.lower() in ("yes", "y", "نعم"):
-            def save_application_to_weaviate(app):
-                print(f"Saving application to Weaviate: {app}")
-                try:
-                    apps = weaviate_client.collections.get("Application")
-                    apps.data.insert(app)
-                    return True
-                except Exception as e:
-                    print(f"Error saving application: {e}")
-                    return False
-            app = {
-                "applicationId": str(uuid.uuid4()), "jobId": session["data"].get("targetJobId"),
-                "applicantName": session["data"].get("applicantName"), "applicantEmail": session["data"].get("applicantEmail"),
-                "coverLetter": session["data"].get("coverLetter", ""), "cvText": session["data"].get("cvText", ""),
-                "skills": session["data"].get("cvSkills", []), "createdAt": get_rfc3339_time()
-            }
-            ok = save_application_to_weaviate(app)
-            session = initial_session()
-            return ("🎉 Your application has been submitted successfully." if ok 
-                    else "⚠️ Failed to save application."), session, False
+        response = chatbot_instance.ask(message)
+        
+        if response.startswith("STARTING_APPLICATION_PROCESS:"):
+            job_id = response.split(":")[1]
+            app_state["mode"] = "APPLYING_CV"
+            app_state["job_id"] = job_id
+            bot_message = f"Starting application for job **{job_id}**. Please upload your CV."
+            history.append((None, bot_message))
+            return history, app_state, gr.update(visible=True)
         else:
-            session = initial_session()
-            return "Application cancelled.", session, False
-
-    # ========== TEAM FLOW ==========
-    if st == "team_action":
-        low = text.lower()
-        if "create" in low or "إنشاء" in low: session["state"] = "team_create_name"; session["data"] = {}; return "Great — what's the team name?", session, False
-        if "join" in low or "انضم" in low: session["state"] = "team_join_name"; session["data"] = {}; return "Okay — what's the name of the team you want to join?", session, False
-        return "Please say 'create' or 'join'.", session, False
-    if st == "team_create_name":
-        session["data"]["team_name"] = text; session["state"] = "team_create_owner"
-        return "Team name saved. Who is the team owner (your name)?", session, False
-    if st == "team_create_owner":
-        session["data"]["owner"] = text; session["state"] = "team_create_skills"
-        return "Owner saved. Please list the team's skills (comma-separated).", session, False
-    if st == "team_create_skills":
-        session["data"]["skills"] = [s.strip() for s in text.split(",") if s.strip()]; session["state"] = "team_create_idea"
-        return "Skills saved. Please write a short idea/description for the project.", session, False
-    if st == "team_create_idea":
-        session["data"]["idea"] = text
-        def save_team_to_weaviate(props):
-            print(f"Saving team to Weaviate: {props}")
-            try:
-                teams = weaviate_client.collections.get("Team")
-                teams.data.insert(props)
-                return True
-            except Exception as e:
-                print(f"Error saving team: {e}")
-                return False
-        team_props = {
-            "teamId": str(uuid.uuid4()), "name": session["data"].get("team_name"),
-            "projectId": None, "members": [session["data"].get("owner")],
-            "skills": session["data"].get("skills", []), "creatorId": session["data"].get("owner"),
-            "createdAt": get_rfc3339_time(), "idea": session["data"].get("idea", "")
-        }
-        saved = save_team_to_weaviate(team_props)
-        session = initial_session()
-        return (f"🎉 Team '{team_props['name']}' created!" if saved else "⚠️ Failed to create team."), session, False
-    if st == "team_join_name":
-        session["data"]["team_name"] = text; session["state"] = "team_join_member"
-        return "What's your name (to add you to the team)?", session, False
-    if st == "team_join_member":
-        session["data"]["member_name"] = text; session["state"] = "team_join_skills"
-        return "Enter your skills (comma-separated).", session, False
-    if st == "team_join_skills":
-        skills = [s.strip() for s in text.split(",") if s.strip()]
-        def update_team_add_member(team_name, member_name, skills):
-            print(f"Adding member {member_name} to team {team_name}")
-            return f"🎉 You have been added to the team '{team_name}'."
-        resp = update_team_add_member(session["data"].get("team_name"), session["data"].get("member_name"), skills)
-        session = initial_session()
-        return resp, session, False
-
-    # ========== RECOMMEND FLOW ==========
-    if st == "recommend_wait_cv":
-        return "Please upload your CV (use the Upload button).", session, True
-
-    return "Sorry — I didn't understand that.", session, False
-
-# ================================
-# Part 4 — Gradio Chat UI wiring
-# ================================
-
-import atexit
-atexit.register(lambda: weaviate_client.close())
-
-def create_initial_session_for_state():
-    return initial_session()
-
-def append_to_history(history: List[Dict[str, str]], role: str, content: str) -> List[Dict[str, str]]:
-    history = history or []
-    history.append({"role": role, "content": content})
-    return history
-
-with gr.Blocks(css="""
-.chatbot {height: 520px; overflow: auto;}
-.user-bubble {background-color: #DCF8C6; padding: 10px; border-radius: 12px; max-width: 75%; float: right; clear: both; margin: 5px; word-wrap: break-word;}
-.bot-bubble {background-color: #F1F0F0; padding: 10px; border-radius: 12px; max-width: 75%; float: left; clear: both; margin: 5px; word-wrap: break-word;}
-.chatbox-container {display: flex; gap: 8px; margin-top: 10px;}
-""") as demo:
-    gr.Markdown("# 💬 EduNatives — Conversational Job Portal")
-    chat_html = gr.HTML(format_chat_html([]))
-    with gr.Row(elem_classes="chatbox-container"):
-        user_input = gr.Textbox(placeholder="Type your message here (e.g. 'apply', 'create team', 'recommend')", lines=2)
-        send_btn = gr.Button("Send", variant="primary")
-    with gr.Row(visible=False) as file_row:
-        cv_uploader = gr.File(label="Upload CV (.pdf/.docx/.txt)", file_count="single", file_types=[".pdf", ".docx", ".txt"], visible=False)
-        upload_btn = gr.Button("Upload CV", visible=False)
-    with gr.Row():
-        clear_btn = gr.Button("Reset Conversation")
-        instructions = gr.Markdown("Commands: apply, create team, join team, recommend — or just ask a question!")
-
-    chat_history_state = gr.State([])        
-    session_state = gr.State(create_initial_session_for_state())
+            history.append((None, response))
+            return history, app_state, gr.update(visible=False)
+            
+    return history, app_state, gr.update(visible=False)
 
-    def handle_send(message: str, history: List[Dict[str, str]], session: dict):
-        history = history or []
-        session = session or initial_session()
-        if message and message.strip():
-            history = append_to_history(history, "user", message.strip())
-        bot_reply, new_session, show_uploader = handle_user_message(session, message or "", uploaded_file=None)
-        history = append_to_history(history, "assistant", bot_reply or "…")
-        html = format_chat_html(history)
-        return "", html, history, new_session, gr.update(visible=show_uploader), gr.update(visible=show_uploader)
 
-    def handle_upload(file_obj, history: List[Dict[str, str]], session: dict):
-        history = history or []
-        session = session or initial_session()
-        filename = getattr(file_obj, "name", "uploaded_file")
-        history = append_to_history(history, "user", f"📎 Uploaded file: {os.path.basename(filename)}")
-        bot_reply, new_session, show_uploader = handle_user_message(session, "", uploaded_file=file_obj)
-        history = append_to_history(history, "assistant", bot_reply or "…")
-        html = format_chat_html(history)
-        return html, history, new_session, gr.update(visible=show_uploader), gr.update(visible=show_uploader)
+with gr.Blocks(theme=gr.themes.Soft(), title="EduNatives Assistant") as demo:
+    
+    application_state = gr.State({"mode": "GENERAL", "job_id": None, "cv_content": None, "cover_letter_content": None})
+    file_uploader = gr.File(label="Upload Document", file_types=['.pdf', '.docx', '.txt'], visible=False)
+    
+    gr.Markdown(
+        """
+        # EduNatives Assistant
+        Ask me anything about jobs, projects, or student availability. I can also help you navigate the EduNatives app.
+        """
+    )
+    
+    chatbot_window = gr.Chatbot(height=450, label="Chat Window", bubble_full_width=False)
+
+    with gr.Column() as examples_container:
+        examples_list = [
+            "What jobs are available at Google?",
+            "Find students with experience in Python and Machine Learning.",
+            "Tell me about the 'AI-Powered Medical Imaging Analysis' project.",
+            "كيف يمكنني كطالب التسجيل في التطبيق؟",
+            "I'm a company, how can I post an internship?"
+        ]
+        with gr.Row():
+            btn1 = gr.Button(examples_list[0], variant='secondary')
+            btn2 = gr.Button(examples_list[1], variant='secondary')
+            btn3 = gr.Button(examples_list[2], variant='secondary')
+        with gr.Row():
+            btn4 = gr.Button(examples_list[3], variant='secondary')
+            btn5 = gr.Button(examples_list[4], variant='secondary')
+        
+        example_buttons = [btn1, btn2, btn3, btn4, btn5]
 
-    def handle_reset():
-        new_hist = []
-        new_session = initial_session()
-        html = format_chat_html(new_hist)
-        return html, new_hist, new_session, gr.update(visible=False), gr.update(visible=False)
+    with gr.Row() as main_input_row:
+        text_input = gr.Textbox(placeholder="Ask your question here...", container=False, scale=7)
+        submit_btn = gr.Button("Send", variant="primary", scale=1)
 
-    send_btn.click(
-        fn=handle_send,
-        inputs=[user_input, chat_history_state, session_state],
-        outputs=[user_input, chat_html, chat_history_state, session_state, cv_uploader, upload_btn],
-        queue=True
+    outputs_list = [chatbot_window, application_state, file_uploader]
+    
+    submit_btn.click(
+        fn=chat_interface_func,
+        inputs=[text_input, chatbot_window, application_state, file_uploader],
+        outputs=outputs_list
     )
-    upload_btn.click(
-        fn=handle_upload,
-        inputs=[cv_uploader, chat_history_state, session_state],
-        outputs=[chat_html, chat_history_state, session_state, cv_uploader, upload_btn],
-        queue=True
+    text_input.submit(
+        fn=chat_interface_func,
+        inputs=[text_input, chatbot_window, application_state, file_uploader],
+        outputs=outputs_list
     )
-    clear_btn.click(
-        fn=handle_reset,
-        inputs=[],
-        outputs=[chat_html, chat_history_state, session_state, cv_uploader, upload_btn],
-        queue=False
+    
+    for btn in example_buttons:
+        btn.click(
+            fn=chat_interface_func,
+            inputs=[btn, chatbot_window, application_state, file_uploader],
+            outputs=outputs_list
+        )
+
+    file_uploader.upload(
+        fn=chat_interface_func,
+        inputs=[gr.Textbox(value="", visible=False), chatbot_window, application_state, file_uploader],
+        outputs=outputs_list
     )
 
+    submit_btn.click(lambda: "", outputs=text_input)
+    text_input.submit(lambda: "", outputs=text_input)
+
+
 if __name__ == "__main__":
     demo.launch(debug=True)