Local-first Edge AI Assistant
- 10 Devlogs
- 32 Total hours
Privacy-first, on-device conversational assistant and research playground for local language models; implemented in Rust and exposed via a lightweight HTTP API.
Privacy-first, on-device conversational assistant and research playground for local language models; implemented in Rust and exposed via a lightweight HTTP API.
I updated the README.md with refreshed information about the project and instructions on how to use/build it. I began work on an interactive REPL to test expression evaluation. My next steps are to give the agent access to this library I developed rather than a REPL for the user, and make it be able to use it with tool calls to evaluate math problems using natural language.
removed assistant indicator and posted backend on github. basically done, project is really basic but was a fun starting point
updated the loader and added an indicator for which assistant is being used
added deep search & framework for pulling images from the web
Working on memory system & multiple tools at once. Basic web search now works
Assistant now shows what tools it’s using
Images with the chatbot responses
chat works now
Working on the chat UI (icons and text are placeholder)
Made the basic window for the app interface