SRS Generator is a collaborative platform designed to streamline the first phase of the Software Development Life Cycle (SDLC) — Software Requirement Specification (SRS) generation.
It enables clients, managers, and developers to work together on a single platform to generate, validate, clarify, version, and track SRS documents efficiently using LLMs and RAG-based intelligence.
- Generate SRS documents using Agent 1 (LLM-powered).
- Receive automated validation from Agent 2, ensuring all functional and non-functional requirements are covered.
- Get suggestions, improvements, and clarification prompts for incomplete or ambiguous requirements.
- View project progress updates shared by the manager.
- Track SRS versions and understand cost implications of feature changes.
- Agent 1: Generates SRS based on client inputs using LLMs.
- Agent 2: Reviews SRS completeness, checks requirement coverage, and provides feedback and suggestions.
- Uses Retrieval-Augmented Generation (RAG) to query the SRS directly.
- Developers can ask natural language questions like:
- “What color theme is expected?”
- “What contact details should be added in the footer?”
- Ensures developers always work with the latest and correct version of the SRS, reducing miscommunication.
- Update and track project progress.
- Monitor SRS versions and feature changes.
- Help optimize project cost by managing feature scope and versions.
- Ensure transparent communication between clients and developers.
- Maintains multiple versions of SRS documents.
- Tracks feature additions, removals, and changes.
- Supports cost optimization by mapping features to scope and impact.
- Frontend: Next.js
- Backend: Node.js, Express.js
- Database: MongoDB
- Authentication: NextAuth
- LLM Provider: Groq
- Embeddings: Hugging Face
- Vector Database: Quadrant (Dockerized)
- RAG Pipeline: Context-aware question answering over SRS
- Containerization: Docker
First, run the development server:
npm install
docker compose up -d
npm run dev