Enterprise R&D Multi-Agent System (rnd_multi_agent)

An enterprise-grade, multi-agent R&D pipeline powered by CrewAI and NeevCloud for automated research, analysis, and report generation.

This repository encompasses the full stack of the R&D Multi-Agent System, consisting of the core agentic workflow, a backend API, and a modern frontend interface.

🏗️ System Architecture

The project is divided into three main components:

1. Core Engine (/rnd_system)

The heart of the application. It uses a hierarchical-sequential hybrid approach where an Orchestrator manages specialist agents (Market Analyst, Tech Researcher, Competitor Intel, Legal, Financial, etc.) who work in parallel to gather, synthesize, and compile extensive research reports.

• Powered by: CrewAI, NeevCloud (minimax-m2.7-highspeed), SearchApi.io

2. Backend API (/api)

A FastAPI wrapper around the core engine. It exposes RESTful endpoints for the frontend to trigger research workflows, query statuses, and retrieve finalized reports and insights.

• Powered by: FastAPI, Python, Uvicorn

3. Frontend Interface (/rnd-system_frontend)

A modern web application that allows users to interact with the multi-agent system. Users can submit research briefs, monitor the parallel execution of specialist agents in real-time, and download finalized multi-format documents (Executive Summaries, Slides, Full Reports).

• Powered by: Next.js, React, TailwindCSS

---

🚀 Quick Start

To run the full stack locally, you will need to start all three components.

1. Start the Backend API

cd api
# Create and activate a virtual environment
py -3.11 -m venv venv
.\venv\Scripts\activate
# Install dependencies
pip install -r requirements.txt
# Run the API server
uvicorn main:app --reload

2. Configure the Core System

Ensure you have the correct API keys in rnd_system/.env and api/.env:

NEEV_API_KEY=your_neev_key
SEARCHAPI_API_KEY=your_searchapi_key

3. Start the Frontend

cd rnd-system_frontend
# Install Node dependencies
npm install
# Run the development server
npm run dev

Navigate to http://localhost:3000 to access the UI.

---

🤖 Agent Roles in the Core System

Agent	Responsibility
Orchestrator	Task decomposition & delegation
Market Analyst	Trends, size, & growth potential
Tech Researcher	Academic research & breakthroughs
Competitor Intel	Market positioning & competitor SWOT
Legal & Risk	Compliance, IP issues, & strategic risk
Financial Modeller	ROI, projections, & cost-benefit
Data Analyst	CSV analysis & statistical insights
Synthesis Agent	Data merging & conflict resolution
Quality Gate	Fact-checking & confidence scoring
Report Writer	Multi-format document generation

---

📁 Repository Structure

/
├── api/                   # FastAPI backend services
├── rnd-system_frontend/   # Next.js web application
└── rnd_system/            # CrewAI multi-agent core logic