Create app.py

app.py

@@ -0,0 +1,112 @@
+import os
+import sys
+import uuid
+import chromadb
+import pandas as pd
+from fastapi import FastAPI, Request
+from fastapi.responses import HTMLResponse
+from fastapi.templating import Jinja2Templates
+from langchain_groq import ChatGroq
+from langchain_community.document_loaders import WebBaseLoader
+from langchain_core.prompts import PromptTemplate
+from langchain_core.output_parsers import JsonOutputParser
+from langchain_core.output_parsers import StrOutputParser
+
+# FastAPI app initialization
+app = FastAPI()
+
+# Get API key from environment variables
+GROQ_API_KEY = os.environ.get("GROQ_API_KEY")
+
+# Set up templates for rendering HTML
+templates = Jinja2Templates(directory="templates")
+
+# --- Initialize Vector Database on Startup ---
+try:
+    df = pd.read_csv("my_portfolio.csv")
+except FileNotFoundError:
+    print("❌ Error: my_portfolio.csv not found.")
+    sys.exit(1)
+
+client = chromadb.PersistentClient('vectorstore')
+collection = client.get_or_create_collection(name="portfolio")
+
+if collection.count() != len(df):
+    if collection.count() > 0:
+        collection.delete(ids=collection.get()['ids'])
+    
+    for _, row in df.iterrows():
+        collection.add(
+            documents=row["Techstack"],
+            metadatas={"links": row["Links"]},
+            ids=[str(uuid.uuid4())]
+        )
+    print("✅ Vector database populated with portfolio data.")
+else:
+    print("✅ Vector database already exists.")
+
+@app.get("/", response_class=HTMLResponse)
+async def get_index(request: Request):
+    """Serves the main robot UI HTML page."""
+    return templates.TemplateResponse("robot_ui.html", {"request": request})
+
+@app.post("/generate")
+async def generate_content(request: Request):
+    form_data = await request.form()
+    job_url = form_data.get('job_url')
+    
+    if not GROQ_API_KEY:
+        return "❌ Error: Groq API key is not set. Please add it to your Space secrets.", 500
+
+    if not job_url:
+        return "Please provide a job URL.", 400
+
+    # --- 1. Validate Groq API Key ---
+    try:
+        llm = ChatGroq(
+            temperature=0,
+            groq_api_key=GROQ_API_KEY,
+            model_name="llama3-70b-8192"
+        )
+        llm.invoke("Test LLM connection.")
+    except Exception as e:
+        return f"❌ Error: Invalid Groq API key or model unavailable. Details: {e}", 500
+
+    # --- 2. Scrape and Extract Job Information ---
+    try:
+        loader = WebBaseLoader(job_url)
+        page_data = loader.load().pop().page_content
+    except Exception as e:
+        return f"❌ Error scraping URL. Please check the URL. Error: {e}", 500
+
+    prompt_extract = PromptTemplate.from_template(
+        """
+        ### SCRAPED TEXT FROM WEBSITE: {page_data}
+        ### INSTRUCTION: Extract the job posting details and return them in JSON format with keys: `role`, `experience`, `skills` and `description`. Only return the valid JSON.
+        ### VALID JSON (NO PREAMBLE):"""
+    )
+    json_parser = JsonOutputParser()
+    chain_extract = prompt_extract | llm | json_parser
+    job = chain_extract.invoke(input={'page_data': page_data})
+    
+    # --- 3. Find Relevant Portfolio Links ---
+    job_skills = job.get('skills', [])
+    relevant_links = collection.query(query_texts=job_skills, n_results=2).get('metadatas', [])
+
+    # --- 4. Generate Cold Email ---
+    prompt_email = PromptTemplate.from_template(
+        """### JOB DESCRIPTION: {job_description}
+        ### INSTRUCTION: You are Mohan, a business development executive at AtliQ. Write a cold email to the client, describing AtliQ's capabilities in fulfilling their needs. Also add the most relevant ones from the following links to showcase Atliq's portfolio: {link_list}
+        ### EMAIL (NO PREAMBLE):"""
+    )
+    chain_email = prompt_email | llm | StrOutputParser()
+    email_content = chain_email.invoke({
+        "job_description": str(job), 
+        "link_list": relevant_links
+    })
+
+    return email_content
+
+if __name__ == '__main__':
+    import uvicorn
+    uvicorn.run("app:app", host="0.0.0.0", port=int(os.environ.get('PORT', 7860)))