minimal_chatbot
Coming SoonView Code
Overview
`minimal_chatbot` is a lightweight full-stack AI chat app. The backend talks to Ollama, while the frontend focuses on a clean chat experience with room for local-first experimentation.Why I built it
I wanted a very small AI chat stack that I could understand completely from top to bottom. Instead of depending on a large hosted platform, this project is centered on local model access through Ollama and a UI that stays intentionally minimal.
What I built
- A FastAPI service for chat, health checks, and NDJSON streaming.
- A Svelte frontend with chat state management and message handling.
- Hooks for voice input and streamed model output.
Technical details
- FastAPI backend exposing
`/chat`,`/stream`, and health routes. - Ollama integration for local model inference.
- Svelte frontend with composables for chat and speech recognition.
- A login-aware interface flow instead of a bare textarea demo.
What I learned
The most useful part of this project is how clearly it separates concerns. Model access, streaming, UI state, and input handling are all small enough to reason about independently, which makes it a good base for later experimentation.
Current status
The project keeps the architecture small on purpose. It is a good foundation for testing interface ideas around local model interaction.
>Newsletter
Stay Updated
Get notified when I publish new articles, tutorials, and project updates. Subscribe for insights and actionable content.