AI RAG
Upload your documents, search semantically, and chat with your knowledge base. Retrieval-Augmented Generation made simple.
From documents to answers in seconds
Your files go through an automated pipeline — parsed, chunked, embedded, and indexed — so AI can retrieve the right context every time.
Supports PDF, DOCX, PPTX, XLSX, HTML, Markdown, TXT, CSV, JSON, EML, and more.
Everything you need for knowledge retrieval
A complete RAG platform — from document ingestion to AI-powered answers.
Document Management
Upload and manage documents in datasets. Automatic parsing with MarkItDown converts PDF, DOCX, PPTX, and 10+ formats to searchable text.
Hybrid Search
Three retrieval modes — semantic (vector), keyword (BM25), and hybrid. Weighted fusion finds the most relevant results every time.
AI Chat
Chat with your knowledge base. AI retrieves relevant context and generates grounded answers with streaming output support.
Multi-Model Support
Connect OpenAI, Anthropic, Google Gemini, Ollama, or any OpenAI-compatible endpoint. Switch models without changing code.
Multi-Tenant Workspaces
Isolated workspaces with role-based access. Owner, Admin, Member, and Viewer roles for secure team collaboration.
Observability
Built-in Langfuse integration for LLM call tracing. Monitor latency, token usage, and retrieval quality in real time.
Dual API architecture
Console API for your dashboard. Service API for programmatic access.
Search your knowledge base
curl -X POST https://api.litestartup.com/api/v1/search/ \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"query": "How do I reset my password?",
"dataset_ids": ["ds-abc123"],
"mode": "hybrid",
"top_k": 5
}'Chat with context
curl -X POST https://api.litestartup.com/api/v1/chats/ \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"message": "Summarize our refund policy",
"conversation_id": "conv-xyz789",
"stream": true
}'Built on production-grade infrastructure
Modern stack designed for reliability and scale.
FastAPI
Backend framework
PostgreSQL
Relational database
Milvus
Vector database
Celery + Redis
Async task queue
React 19
Frontend UI
Tailwind CSS
Styling
Langfuse
LLM observability
Docker
Container deployment
Up and running in 3 steps
Connect your model
Add your LLM provider — OpenAI, Anthropic, Gemini, or Ollama. Configure your embedding and inference models.
Upload your documents
Create a dataset and upload files or paste text. Documents are parsed, chunked, and indexed automatically in the background.
Search and chat
Use semantic search to find relevant content, or start a conversation and let AI answer questions grounded in your data.
Built for real-world use cases
Internal Knowledge Base
Upload company docs, policies, and playbooks. Let employees search and get instant answers instead of digging through wikis.
Customer Support Bot
Feed your help center articles and FAQs into a dataset. Build a support chatbot that answers questions with accurate, sourced responses.
Research Assistant
Upload research papers, reports, and notes. Ask questions across your entire research library and get cited answers.
Product Documentation
Index your API docs, changelogs, and guides. Developers can ask natural language questions and get code-relevant answers.
Ready to build your knowledge base?
Start with the free tier. Upload your first documents in minutes.
Start Free Trial