<<<<<<< HEAD

ARAA — Autonomous Research Assistant AI

ARAA (Autonomous Research Assistant AI) is an advanced agentic AI project designed to automate complex research workflows. It leverages LLM-based agents to plan research goals, break them into actionable subtasks, perform autonomous searches, and summarize information efficiently. It mimics human decision-making and self-directed behavior, enabling researchers to focus on strategic work.

ARAA is built using the LLaMA 3 model via Ollama, integrated with LangChain for agent orchestration and task execution.

---

Features

Feature	Description
Goal Planning	Breaks high-level research goals into sequential, prioritized tasks using Planner Agent
Autonomous Search	Searches the web autonomously using DuckDuckGo and keyword expansion via Searcher Agent
Summarization	Summarizes results into concise bullet points, pros/cons, or key insights
Agentic Workflow	Simulates human decision-making to select and execute tasks effectively
Extensible	Modular design allows integration of new agents or tools
Frontend Integration Ready	Supports easy integration with a React-based frontend for interactive dashboards

---

Folder Structure

ARAA/
├── venv/                     # Python virtual environment
├── agents/                   # AI agent modules
│   ├── planner_agent.py      # Plans research goals and creates task sequences
│   └── searcher_agent.py     # Performs automated searches and summarizes information
├── main.py                   # Main entry point for running ARAA
├── requirements.txt          # Python dependencies
└── .gitignore

---

Setup Instructions

1️⃣ Clone the Repository

git clone https://github.com/Mayu-infinite/ARAA.git
cd ARAA

2️⃣ Create and Activate Python Virtual Environment

python3 -m venv venv
source venv/bin/activate

3️⃣ Install Required Python Packages

pip install --upgrade pip
pip install langchain langchain-ollama langchain-community duckduckgo-search openai pandas numpy

Optional for image processing or advanced tasks:

pip install opencv-python pillow scikit-learn

4️⃣ Start Ollama Server

In a separate terminal:

ollama serve

Wait until you see:

Listening on 127.0.0.1:11434

5️⃣ Run ARAA

With the virtual environment active:

python main.py

6️⃣ Interacting with ARAA

1. Enter your research goal when prompted.
2. ARAA generates a numbered task plan with subtasks.
3. Select the task number to execute; ARAA performs search, summarization, and outputs results.
4. Repeat for subsequent tasks to progress toward the overall research goal.

---

Example Usage

Input	Description	Output
"Build a research plan for AI-based autonomous drones"	User provides overall research goal	Numbered subtask plan
Select task 2	Execute a subtask	Summarized search results with links and key points

---

GitHub Setup

To version control your project and push updates:

git add .
git commit -m "Initial commit: Complete ARAA setup with planner & searcher agents"
git push -u origin main

---

Future Enhancements

Enhancement	Description
Additional Agents	Data analysis, visualization, scraping agents can be added
Frontend Dashboard	React frontend for interactive task visualization and summaries
Web Scraping & PDF Analysis	Automate extraction from academic papers, reports, and online resources
Enhanced Agentic Reasoning	Implement task prioritization and chaining for better decision-making
Multi-User Support	Collaborative research environment with multiple users and profiles

---

System Requirements

• OS: Linux (Fedora recommended) or macOS/Windows
• Python: >=3.11
• RAM: At least 8GB
• Ollama: Installed and running
• Internet: Required for autonomous search and updates

---

Author

Mayuri and Prasangeet

License

This project is open-source under the MIT License.

---

Contact & Support

• GitHub Issues: https://github.com/Mayu-infinite/ARAA/issues
• Email for collaboration and questions: (your GitHub email)

---

This README provides a professional, comprehensive overview of ARAA, including features, setup, usage, examples, future plans, and contact info for a GitHub-ready project page.

ARAA - Autonoumous Reasearch and Action Agent

ARAA is an agentic AI system designed to think, plan, and act autonomously — similar in spirit to Gemini or GPT-based agent frameworks — but built entirely using free, open-source tools like LangChain, local LLMs (via Ollama/llama.cpp), and ChromaDB for vector memory.

2994e7a (updated project)