🧠 Multi-Agent AI Researcher

A smart, interactive research tool that helps you explore HackerNews like never before — using the power of AI agents powered by GPT-4o.

🔍 What Is This?

This is a Streamlit-powered app that uses multiple AI assistants working together to help you:

• Discover top stories on HackerNews
• Research active users
• Generate insightful blog posts, reports, or social media updates based on what you're curious about

Think of it as a team of expert researchers in your pocket — all running on AI.

---

🚀 How to Get Started

1. Clone the Repository

Open your terminal and run:

git clone https://github.com/anandsuraj/multi_agent_researcher.git
cd multi_agent_researcher

2. Install Dependencies

Make sure you have Python installed (preferably 3.9+), then install the required packages:

pip install -r requirements.txt

3. Get an OpenAI API Key

You'll need access to OpenAI's GPT-4 model. Follow these steps:

• Sign up at OpenAI Platform
• Create an API key from your dashboard
• Keep this key handy — you’ll enter it when you start the app

💡 Pro tip: Never share or commit your API key publicly.

4. Launch the App

Run the app with:

streamlit run research_agent.py

Or if you're using a virtual environment:

./venv/bin/python -m streamlit run research_agent.py

Then open the link shown in your browser.

---

🤖 How It Works

When you launch the app:

1. You’ll be asked to enter your OpenAI API key

2. Behind the scenes, three AI agents are created:

   • Story Researcher: Specializes in digging into HackerNews stories
   • User Researcher: Focuses on user profiles, activity, and article summaries
   • HN Assistant: Your main assistant who coordinates between the two researchers

3. Type in your research question or topic (e.g., "What are the most popular AI stories this week?" or "Tell me about user 'pg' on HackerNews.")

4. The HN Assistant takes over:

   • Decides which agent should do what
   • Fetches data from HackerNews
   • Uses GPT-4o to generate a well-written summary or content piece

5. You get a clear, well-structured output, ready to use for blogging, reporting, or sharing!

---

📌 Example Use Cases

Here are some things you can ask:

• "Summarize the top 5 stories from HackerNews today."
• "Who is the user 'dhh' on HackerNews?"
• "Write a Twitter/X thread about the latest trends in AI from HackerNews."
• "Generate a blog post summarizing the best productivity advice from recent HackerNews threads."

---

🛠️ Built With

• Streamlit: For the web interface
• OpenAI GPT-4o: To power the AI agents
• LangChain: For managing tools and agent workflows
• Python: Core programming language

---

✅ Future Enhancements (Ideas)

• Support for more platforms beyond HackerNews
• Export generated content to PDF or Markdown
• Save and reuse previous research queries
• Add support for other LLM providers (like Anthropic, Gemini, etc.)

---

🤝 Contributing

Contributions are welcome! If you'd like to improve the app, fix bugs, or add features, feel free to open a PR or issue on GitHub.

---

📬 Feedback & Questions?

If you have any questions, suggestions, or want to collaborate, reach out to the repo owner or open an issue on GitHub.

---

Enjoy researching smarter with AI! 🚀
Let the agents do the heavy lifting while you focus on asking great questions.

---