Intramind.ai is an enterprise-grade AI-powered knowledge platform that helps organizations unlock information from internal documents through intelligent conversations.
Built with a high-performance RAG (Retrieval-Augmented Generation) Pipeline, it allows teams to query company knowledge naturally and receive context-aware answers with citations.
frontend/
├── app/ # Next.js App Router
│ ├── (auth)/ # Auth routes (login, signup)
│ └── dashboard/ # Main app dashboard (workspace, chat, docs)
├── components/ # Reusable UI components
├── hooks/ # Custom React hooks
├── services/ # API service layer
├── store/ # State management (Zustand)
└── utils/ # Utility functions
backend/
├── api/ # API routes & endpoints
├── core/ # App configuration & security
├── models/ # Database models (SQLAlchemy/Tortoise)
├── services/ # Business logic layer
├── ai/ # RAG pipeline & AI logic
├── embeddings/ # Text embedding logic
├── vectorstore/ # FAISS / Pinecone integration
└── db/ # Database session & migrations
infra/ # Docker & K8s configurations
.github/workflows/ # CI/CD pipelines (Lint, Test, Build, Deploy)
- Document Upload: Multi-format support (PDF, MD, TXT).
- Text Extraction: High-accuracy parsing.
- Chunking: Intelligent text splitting.
- Embedding: Converting text to high-dimensional vectors.
- Vector Store: Storage in FAISS/Pinecone.
- Query: Natural language retrieval.
- Similarity Search: Finding the most relevant context.
- LLM Response: AI-generated answers with source citations.
- Auth: JWT + refresh tokens.
- Access Control: RBAC enforcement.
- Protection: Rate limiting (Redis) & Input sanitization.
- Network: HTTPS everywhere & Secure headers.