app.py

import os
import gradio as gr
import numpy as np
import faiss
from sentence_transformers import SentenceTransformer
from groq import Groq

# ---------------- CONFIG ----------------
GROQ_API_KEY = os.getenv("GROQ_API_KEY")
if not GROQ_API_KEY:
    raise ValueError("Please set GROQ_API_KEY in Hugging Face Secrets")

# ---------------- INIT ----------------
client = Groq(api_key=GROQ_API_KEY)
embedder = SentenceTransformer("all-MiniLM-L6-v2")

EMBED_DIM = 384
faiss_index = faiss.IndexFlatIP(EMBED_DIM)
candidates = []

# ---------------- HELPERS ----------------
def embed_text(text: str):
    vec = embedder.encode([text])
    vec = vec / np.linalg.norm(vec)
    return vec.astype("float32")

def llm_generate(prompt: str):
    completion = client.chat.completions.create(
        model="llama-3.1-8b-instant",
        messages=[{"role": "user", "content": prompt}],
        temperature=0.3
    )
    return completion.choices[0].message.content

# ---------------- CORE FEATURES ----------------
def add_candidate(name, skills, experience):
    if not name or not skills:
        return (
            "⚠️ Please fill in required fields.",
            gr.update(),
            gr.update(),
            gr.update()
        )

    profile = f"Name: {name}\nSkills: {skills}\nExperience: {experience}"
    vec = embed_text(profile)
    faiss_index.add(vec)
    candidates.append(profile)

    return (
        f"✅ Candidate **{name}** added successfully.",
        gr.update(value=""),   # clear name
        gr.update(value=""),   # clear skills
        gr.update(value="")    # clear experience
    )


def generate_bio(raw_data):
    prompt = f"""
You are a professional HR recruiter.
Write a concise, formal candidate bio.

Rules:
- Professional tone
- No emojis
- No exaggeration
- Max 120 words

Candidate Data:
{raw_data}
"""
    return llm_generate(prompt)

def rewrite_job(job_desc):
    prompt = f"""
Rewrite the job description below to meet modern recruitment standards.
Make it inclusive, clear, and well-structured.

Job Description:
{job_desc}
"""
    return llm_generate(prompt)

def recommend_candidates(job_query):
    if not candidates:
        return "No candidates available.", "", "⚠️ Add candidates first."

    job_vec = embed_text(job_query)
    scores, ids = faiss_index.search(job_vec, k=min(3, len(candidates)))

    ranked, reasons = [], []

    for i, idx in enumerate(ids[0]):
        candidate = candidates[idx]
        explain_prompt = f"""
Explain why this candidate matches the job.
Focus only on skills and experience.

Candidate:
{candidate}

Job Requirements:
{job_query}
"""
        ranked.append(f"### Rank {i+1}\n{candidate}")
        reasons.append(llm_generate(explain_prompt))

    return "\n\n".join(ranked), "\n\n".join(reasons), "✅ Recommendations generated"

# ---------------- LINKEDIN POST GENERATOR (NEW) ----------------
def generate_linkedin_post(context_type, content):
    prompt = f"""
You are a LinkedIn content strategist for HR & AI professionals.

Create a LinkedIn post based on the context below.

Context Type: {context_type}

Rules:
- Professional but engaging tone
- Short paragraphs
- Strong hook at the beginning
- Clear CTA at the end
- 5–8 relevant hashtags
- Max 2 emojis
- Do NOT mention AI model names

Content:
{content}
"""
    return llm_generate(prompt)

# ---------------- UI ----------------
theme = gr.themes.Soft(
    primary_hue="blue",
    secondary_hue="slate",
    neutral_hue="gray",
)

with gr.Blocks(theme=theme) as demo:

    gr.Markdown("""
    # 🤖 AI Recruitment Assistant  
    **Recruit smarter. Hire faster. Share better.**
    """)

    gr.Markdown("---")

    with gr.Tabs():

        # -------- Add Candidate --------
        with gr.Tab("➕ Add Candidate"):
            with gr.Row():
                with gr.Column():
                    name = gr.Textbox(label="Candidate Name *")
                    skills = gr.Textbox(label="Key Skills *")
                    experience = gr.Textbox(label="Experience Summary", lines=4)
                    add_btn = gr.Button("Add Candidate", variant="primary")
                    status = gr.Markdown()

                with gr.Column():
                    gr.Markdown("""
                    ### Tips
                    - Mention years of experience  
                    - Use clear skill names  
                    - Avoid generic wording  
                    """)

            add_btn.click(
    add_candidate,
    inputs=[name, skills, experience],
    outputs=[status, name, skills, experience]
)

        # -------- Bio Generator --------
        with gr.Tab("🧾 Candidate Bio Generator"):
            raw_data = gr.Textbox(lines=6, placeholder="Paste raw CV or notes...")
            bio_btn = gr.Button("Generate Bio", variant="primary")
            bio_output = gr.Textbox(lines=6)

            bio_btn.click(generate_bio, raw_data, bio_output)

        # -------- Job Rewriter --------
        with gr.Tab("📄 Job Description Optimizer"):
            job_desc = gr.Textbox(lines=6, placeholder="Paste job description...")
            rewrite_btn = gr.Button("Rewrite Professionally", variant="primary")
            rewrite_output = gr.Textbox(lines=6)

            rewrite_btn.click(rewrite_job, job_desc, rewrite_output)

        # -------- Recommendation --------
        with gr.Tab("🎯 Candidate Recommendation"):
            job_query = gr.Textbox(lines=5, placeholder="Enter job requirements...")
            rec_btn = gr.Button("Find Best Matches", variant="primary")

            with gr.Row():
                rec_candidates = gr.Markdown()
                rec_explanation = gr.Textbox(lines=10)

            rec_status = gr.Markdown()

            rec_btn.click(
                recommend_candidates,
                job_query,
                [rec_candidates, rec_explanation, rec_status]
            )

        # -------- LinkedIn Post Generator (NEW) --------
        with gr.Tab("📣 LinkedIn Post Generator"):
            gr.Markdown("### Turn AI outputs into LinkedIn-ready posts")

            post_type = gr.Dropdown(
                label="Post Context",
                choices=[
                    "Candidate Bio Highlight",
                    "Job Hiring Announcement",
                    "AI-Based Candidate Recommendation"
                ]
            )

            post_input = gr.Textbox(
                lines=8,
                placeholder="Paste generated bio, job description, or recommendation..."
            )

            post_btn = gr.Button("Generate LinkedIn Post", variant="primary")
            post_output = gr.Textbox(lines=10, label="LinkedIn Post")

            post_btn.click(
                generate_linkedin_post,
                inputs=[post_type, post_input],
                outputs=post_output
            )

        # -------- About --------
        with gr.Tab("📘 About"):
            gr.Markdown("""
            ### What is this app?
            An AI-powered recruitment assistant that helps recruiters:
            - Manage candidates
            - Improve hiring content
            - Match talent intelligently
            - Generate social media content

            ### Use Case
            Generate → Optimize → Match → Share on LinkedIn

            ### Ideal for
            Recruiters • HR Tech • AI Portfolios • Startups
            """)

    gr.Markdown("--- Built for Hugging Face Spaces 🚀")

demo.launch()