Full feature set
Built for every AI use case
From quick RAG prototypes to enterprise-grade AI deployments, every feature you need is already here.
AI Servers
- Multi-model support (Gemini, GPT, Claude)
- Per-server configuration & isolation
- System prompt customization
- Conversation history & context
- Token usage tracking
RAG Knowledge Base
- PDF, Word, TXT document upload
- Automatic chunking & embedding
- pgvector semantic search
- Cosine similarity ranking
- Per-server document isolation
Dynamic Tools
- Define REST API tools at runtime
- Automatic Java class generation
- Spring AI @Tool annotation support
- OAuth & API key auth per tool
- Live reload without redeploy
Embeddable Chat
- One-line embed script
- Fully customizable UI
- CORS-safe iframe & widget mode
- Mobile responsive
- Dark & light theme support
Security & Auth
- Per-server API key management
- OAuth 2.0 tool authentication
- Request signing & header injection
- Rate limiting per server
- Audit log ready
Developer Experience
- REST API for all operations
- Swagger/OpenAPI docs
- Docker & docker-compose support
- Spring Boot 3 + Java 21
- Webhook event support (coming soon)
How it works
Three steps from zero to a fully working AI assistant.
01
Create a Server
Define your AI server, pick a model, and write your system prompt.
02
Add Knowledge
Upload your documents. We chunk, embed, and index them automatically.
03
Deploy & Chat
Embed the chat widget or call the API. Your AI is live instantly.