JARVIS Dev is a local AI developer assistant designed to make working with existing codebases faster and easier. Developers often spend significant time understanding unfamiliar projects, finding relevant files, tracing how components connect, and repeatedly searching through code. JARVIS Dev addresses this workflow by loading a project's context and allowing developers to interact with that context through project analysis and code-focused questions. The system runs AI locally using Ollama and Llama 3.1 8B, allowing the developer to work with their project without relying on a cloud-based AI service for the core developer-assistance workflow. JARVIS Dev includes project context loading, project analysis, code Q&A, voice interaction, and a custom HUD interface. The goal is to reduce the friction involved in developer onboarding and ongoing application maintenance. The project was built as a working prototype for the IBM Bob 2.0 hackathon, with a focus on applying AI to a practical developer workflow rather than creating a general-purpose chatbot.
Category tags: