app.py

import gradio as gr
from huggingface_hub import InferenceClient
import os
import gradio as gr
from openai import OpenAI
from dotenv import load_dotenv
from pypdf import PdfReader

load_dotenv(override=True)

api_key = os.getenv("OPENROUTER_API_KEY")
cv_link= os.getenv("CV_LINK")
name = os.getenv("NAME")


if not api_key:
    raise ValueError("OPENROUTER_API_KEY not found in .env file. Please set it.")

client = OpenAI(
    base_url="https://openrouter.ai/api/v1",
    api_key=api_key
)



summary = (
    "I am a 3 years experienced Backend Software Engineer with a strong focus on building scalable, well-architected systems "
    "using NestJS, GraphQL, and PostgreSQL. I have proven experience in developing high-performance APIs, "
    "applying Domain-Driven Design (DDD), and leading AI-powered solutions such as RAG-based chatbots "
    "integrated with vector databases like Qdrant and Pinecone. I am skilled in clean architecture, "
    "microservices, and production-grade e-commerce systems, and I'm adept at bridging business needs with "
    "technical implementation through both code and technical writing. I have contributed to real-time systems "
    "using gRPC, Redis, and message queues (e.g., RabbitMQ). I'm passionate about automation, developer "
    "tools (e.g., n8n, Copilot, Cursor), and mentoring through technical content. As an ALX alumna, I have a "
    "hands-on mindset and a continuous learning attitude. I am also the author of a technical blog on "
    "software engineering and AI systems: shazaali.substack.com."
)

linkedin_text = "My LinkedIn Profile: https://www.linkedin.com/in/shazaali/\n\n"
try:
    reader = PdfReader("linkedin.pdf")
    for page in reader.pages:
        linkedin_text += page.extract_text() + "\n"
except FileNotFoundError:
    print("Warning: 'linkedin.pdf' not found. The bot will rely only on the summary.")
    linkedin_text = "LinkedIn profile data is not available."

system_prompt = (
    f"You are acting as {name}. You are answering questions on my personal website, "
    f"particularly questions related to my career, background, skills, and experience. "
    f"Your responsibility is to represent me as faithfully as possible. "
    f"You are given a summary of my background and my LinkedIn profile to use for answering questions. "
    f"Be professional, friendly, and engaging, as if you are talking to a potential client or future employer. "
    f"If you don't know the answer based on the provided context, it's better to say so than to invent information. "
    f"Always stay in character as {name}. Answer in the same language as the user's question without disclosing any private information."
    f"\n\n## My Summary:\n{summary}\n\n## My LinkedIn Profile Text:\n{linkedin_text}\n\n## My CV:\n{cv_link}"
)


def chat(message, history):
    """
    Handles the chat logic by formatting messages and calling the OpenAI API.
    """

    formatted_history = []
    for user_msg, assistant_msg in history:
        formatted_history.append({"role": "user", "content": user_msg})
        formatted_history.append({"role": "assistant", "content": assistant_msg})

    messages = [
        {"role": "system", "content": system_prompt},
        *formatted_history,
        {"role": "user", "content": message}
    ]

    try:
        response = client.chat.completions.create(
            model="openai/gpt-3.5-turbo",
            max_tokens=300,
            messages=messages,
            temperature=0.7
        )
        return response.choices[0].message.content
    except Exception as e:
        print(f"An error occurred: {e}")
        return "Sorry, I encountered an error while processing your request. Please try again."


interface = gr.ChatInterface(
    fn=chat,
    title=f"Chat with {name}",
    description="Ask me about my experience, skills, or projects",
    examples=[
        ["What are your main technical skills?"],
        ["Tell me about your experience with AI and RAG chatbots."],
        ["¿Hablas español?"]
    ]
)

if __name__ == "__main__":
    interface.launch()