// demo · ai
AI assistant workbench
Chat UI with streaming feel, conversation history, and a prompt template library.
Overview
An LLM chat product. Streaming-style responses (token-by-token render), a conversation sidebar with history, model picker, and a small prompt template library on the admin side.
What's in this demo
- Streaming chat with token-by-token render
- Conversation history sidebar
- Model picker (gpt-style + claude-style)
- Prompt template library (read-only)
Stack we'd ship this on
This demo backs
Notes
Frontend-only. Mock data resets on refresh. Use it to feel the surface — for production we'd wire it to a real backend, auth, payments, and observability.