🔬 AI Research Agents

A multi-agent AI system where specialized agents collaborate to produce comprehensive research reports

Enter any topic → Watch AI agents research, write, and review → Get a polished report.

---

Architecture

┌──────────────────────────────────────────────────────────────┐
│                    AI Research Agents                          │
├──────────────────────────────────────────────────────────────┤
│                                                               │
│   User Input (Topic)                                          │
│        │                                                      │
│        ▼                                                      │
│   ┌─────────────┐    ┌─────────────┐    ┌─────────────┐     │
│   │ 🔍 RESEARCHER│───▶│ ✍️ WRITER   │───▶│ 🔎 REVIEWER │     │
│   │              │    │             │    │             │     │
│   │ Gathers info │    │ Creates     │    │ Reviews &   │     │
│   │ & key facts  │    │ polished    │    │ improves    │     │
│   │              │    │ report      │    │ final       │     │
│   └─────────────┘    └─────────────┘    └─────────────┘     │
│        │                    │                  │              │
│        ▼                    ▼                  ▼              │
│   Research Notes      Draft Report       Final Report        │
│                                                               │
│   ┌──────────────────────────────────────────────────┐       │
│   │              🎯 ORCHESTRATOR                      │       │
│   │  Coordinates pipeline, manages state, callbacks   │       │
│   └──────────────────────────────────────────────────┘       │
│                                                               │
│   Frontend: Streamlit  │  LLM: Claude  │  Lang: Python       │
└──────────────────────────────────────────────────────────────┘

How It Works

Each agent is a specialized AI with its own system prompt and role:

Agent	Role	System Prompt Focus
🔍 Researcher	Gathers comprehensive information	Deep research, key concepts, current state, trends
✍️ Writer	Creates a polished report	Structure, clarity, executive summary, takeaways
🔎 Reviewer	Reviews and improves	Accuracy, completeness, clarity, final improvements

The Orchestrator coordinates the pipeline:

1. Sends topic to Researcher → receives research notes
2. Sends notes to Writer → receives draft report
3. Sends draft to Reviewer → receives final polished report

Features

• 3-Agent Pipeline — Researcher, Writer, Reviewer work sequentially
• Real-Time Pipeline Visualization — Watch each agent activate as it works
• Tabbed Output — View research notes, draft, and final report separately
• Report Export — Download the full report as markdown
• Modular Design — Each agent is independently configurable
• Orchestrator Pattern — Clean pipeline coordination with callbacks
• Docker Support — Containerized deployment

Quick Start

# Clone
git clone https://github.com/Shrinija17/ai-research-agents.git
cd ai-research-agents

# Setup
python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

# API Key
echo "ANTHROPIC_API_KEY=your-key-here" > .env

# Run
streamlit run app.py

Project Structure

ai-research-agents/
├── app.py                      # Streamlit web interface
├── src/
│   ├── config.py               # Configuration & settings
│   ├── agents.py               # Agent definitions & system prompts
│   └── orchestrator.py         # Pipeline coordination
├── tests/
│   └── test_orchestrator.py    # Unit tests
├── Dockerfile                  # Container support
├── requirements.txt            # Dependencies
└── .env.example                # API key template

Running Tests

python -m pytest tests/ -v

Try These Topics

• "Future of AI Agents in 2025"
• "RAG vs Fine-tuning: When to use which"
• "How autonomous vehicles navigate"
• "AI applications in healthcare diagnostics"
• "The rise of edge computing"

---

Built by Shrinija Kummari · Powered by Claude AI