🤖 Multi-Agent Research Assistant

An intelligent Agentic AI Research Assistant built using LangGraph, LangChain, Groq LLM, ChromaDB, Tavily Search, and Wikipedia, capable of autonomously planning research, gathering information from multiple sources, reflecting on the quality of its work, refining queries, and maintaining long-term semantic memory.

🌐 Live Demo

🚀 Try the application here

Streamlit App:
https://multiagentresearchassistant-chethan.streamlit.app/

GitHub Repository:
https://github.com/ChethanMShivappa/MultiAgentResearchAssistant

---

🚀 Features

✅ Multi-Agent Architecture using LangGraph

✅ Intelligent Research Planning

✅ Multi-Source Information Retrieval

• Tavily Search API
• Wikipedia Search

✅ Reflection Agent

• Evaluates research quality
• Scores research output
• Identifies strengths & weaknesses
• Suggests improvements

✅ Query Refinement

• Automatically improves weak queries
• Performs iterative research

✅ Persistent Semantic Memory

• ChromaDB Vector Database
• Retrieves relevant past research
• Context-aware responses

✅ Interactive Streamlit Dashboard

• Modern UI
• Workflow visualization
• Research report
• Memory display
• Sources used
• Analytics

---

🏗 Architecture

                User Query
                     │
                     ▼
           Planner Agent
                     │
                     ▼
         Search Manager Agent
          ┌──────────┴──────────┐
          ▼                     ▼
   Tavily Search        Wikipedia Search
          └──────────┬──────────┘
                     ▼
            Research Agent
                     │
                     ▼
           Reflection Agent
          ┌──────────┴──────────┐
          │                     │
     Approved?              Needs Improvement
          │                     │
          ▼                     ▼
     Memory Agent       Query Refiner Agent
          │                     │
          └──────────┬──────────┘
                     ▼
               Final Response

---

📂 Project Structure

MultiAgentResearchAssistant/
│
├── agents/
│   ├── planner_agent.py
│   ├── research_agent.py
│   ├── reflection_agent.py
│   ├── query_refiner_agent.py
│
├── graph/
│   ├── graph.py
│   ├── state.py
│
├── memory/
│   ├── chroma_memory.py
│
├── prompts/
│
├── schemas/
│
├── tests/
│
├── tools/
│   ├── tavily_search.py
│   ├── wikipedia_search.py
│
├── reports/
│
├── utils/
│
├── app.py
├── config.py
├── requirements.txt
└── README.md

---

⚙ Tech Stack

Category	Technology
LLM	Groq (Llama 3.3 70B)
Framework	LangGraph
LLM Framework	LangChain
Memory	ChromaDB
Embeddings	sentence-transformers
Search	Tavily API
Knowledge Base	Wikipedia
Frontend	Streamlit
Language	Python

---

🧠 Workflow

1. Planner Agent

Creates a structured research plan based on the user's query.

↓

2. Search Manager

Collects information from:

• Tavily Search
• Wikipedia

↓

3. Research Agent

Combines information into a comprehensive research report.

↓

4. Reflection Agent

Evaluates the generated report based on:

• Completeness
• Accuracy
• Coverage
• Clarity

↓

5. Query Refiner

If the report quality is insufficient:

• Improves the research query
• Runs another research cycle

↓

6. Memory Agent

Stores research in ChromaDB and retrieves relevant past knowledge for future queries.

---

📸 Screenshots

Dashboard

Workflow

Research Report

🛠 Installation

Clone Repository

git clone https://github.com/ChethanMShivappa/MultiAgentResearchAssistant.git

cd MultiAgentResearchAssistant

---

Create Virtual Environment

Windows

python -m venv venv

venv\Scripts\activate

Linux / Mac

python3 -m venv venv

source venv/bin/activate

---

Install Dependencies

pip install -r requirements.txt

---

Configure Environment

Create a .env file

GROQ_API_KEY=YOUR_GROQ_API_KEY

TAVILY_API_KEY=YOUR_TAVILY_API_KEY

---

Run Streamlit

streamlit run app.py

---

🧪 Running Tests

python -m tests.test_graph

---

📊 Example Research Flow

User Query

↓

Planner Agent

↓

Search Manager

↓

Research Agent

↓

Reflection Agent

↓

Approved?

↓

Memory Storage

↓

Final Research Report

---

🌟 Key Highlights

• Autonomous Multi-Agent Workflow
• Reflection-Based Quality Improvement
• Long-Term Semantic Memory
• Retrieval-Augmented Research
• Interactive Dashboard
• Modular Agent Design
• Production-Ready Architecture

---

🚀 Future Improvements

• PDF Report Export
• Citation Generation
• Multi-LLM Support
• Local LLM Integration
• Voice Interface
• Research History Database
• Docker Deployment
• Cloud Deployment (AWS / Azure / GCP)

---

👨‍💻 Author

Chethan M S

AI Engineer | Generative AI | Agentic AI | Machine Learning

GitHub: https://github.com/ChethanMShivappa

LinkedIn: linkedin.com/in/chethan-m-s-b544673a9

---

📜 License

This project is licensed under the MIT License.

---

⭐ Support

If you found this project useful, consider giving it a ⭐ on GitHub.

It helps others discover the project and motivates further development.

---

🙏 Acknowledgements

• LangGraph
• LangChain
• Groq
• ChromaDB
• Tavily
• Wikipedia
• Streamlit