Social Profile Analyzer (Agentic AI Project)

This project is an intelligent Agentic AI web application built using LangChain and OpenRouter’s DeepSeek model. It takes a person's name as input and returns a professional profile summary, interesting facts, potential ice-breakers, and topics of interest based on publicly available data from LinkedIn and Twitter.

💡 Features

• 🔍 Name-based Profile Search (LinkedIn & Twitter)
• 📄 Summarized Bio & Facts
• 💬 Conversation Starters & Ice-breakers
• 📚 Topics of Interest
• 🧠 Powered by LangChain + DeepSeek: R1 0528 (free)

---

📦 Tech Stack

• 🦜 LangChain (Agents + Tools)
• 🌐 Flask (Frontend + Backend)
• 🧠 Deepseek R1 LLM
• 🕸️ Tavily Search API
• 🐦 Twitter API (via Tweepy)
• 🔗 Proxycurl (LinkedIn scraping)
• 🧪 LangSmith (Optional for tracing/debugging)

---

🔐 Environment Variables

To run this project locally or on Hugging Face, you must create a .env file in the root directory with the following keys:

OPENROUTER_API_KEY
PROXYCURL_API_KEY
TAVILY_API_KEY

TWITTER_API_KEY
TWITTER_API_KEY_SECRET
TWITTER_BEARER_TOKEN
TWITTER_ACCESS_TOKEN
TWITTER_ACCESS_TOKEN_SECRET

LANGCHAIN_TRACING              # (Optional)
LANGSMITH_ENDPOINT             # Required if tracing enabled
LANGCHAIN_API_KEY              # Required if tracing enabled
LANGCHAIN_PROJECT              # Required if tracing enabled

---

🤖 LLM Configuration

This project uses the deepseek/deepseek-r1-0528:free model via OpenRouter for language generation.

If you want to use a different OpenRouter-compatible LLM, make sure to update the model parameter accordingly:

from langchain_openai import ChatOpenAI
import os

llm = ChatOpenAI(
    model="deepseek/deepseek-r1-0528:free",  
    base_url="https://openrouter.ai/api/v1",
    api_key=os.getenv("OPENROUTER_API_KEY"),
    temperature=0,
)

💵 Note on API Usage

This project uses paid API services:

• Proxycurl for LinkedIn data scraping
  (Includes 10 free credits to start)
• Twitter API (paid) for accessing Twitter data

---

🔄 Behavior with Mock Data

Because these APIs are paid, the deployed version uses mock data.

📝 What does that mean?

• If you enter any name, the app will always return a hardcoded mock profile.
• This allows the demo to work without requiring paid API keys.

---

🔧 How to Enable Real-time Data Scraping

To fetch real LinkedIn and Twitter data, follow these steps:

1. Purchase API keys for Proxycurl and Twitter.
2. Set the keys in your .env file.
3. Modify the file profile_analyzer.py:

# 🔄 Replace this mock logic:
linkedin_data = scrape_linkedin_profile(linkedin_url, mock=True)
tweets = scrape_user_tweets_mock()

# ✅ With this:
linkedin_data = scrape_linkedin_profile(linkedin_url)
tweets = scrape_user_tweets(twitter_url)

---

🙌 Contributing

Contributions are welcome! If you find a bug or have suggestions for improvements, feel free to open an issue or submit a pull request.

---

📬 Contact

For any questions or feedback, reach out to me via LinkedIn or open an issue on this repository.

---

⭐️ Show Your Support

If you like this project, please give it a ❤️ on Huggingface!

---