archiTECH

landing page

Inspiration

With a Computer Engineer on our team, a struggle they often came across was a lack of accurate AI assistance when it came to working with hardware, with them often needing to heavily review the AI code, mitigating the aid that AI should be offering. "Why isn't there a tool like Cursor for hardware?", we thought, which is why we decided to build a tool aimed specifically for people in this field.

What it does

Designed to be a tool to streamline hardware engineering workflows, specifically for Verilog code, with fine tuned LLMs with AI Assistance (including smart autocompletes and agentic modes) to help engineers streamline their coding process. To futher enhance the process, we included automated test bench creations as well as automatic simulations embedded directly into the application.

How we built it

We built this application as a webapp using FastAPI and React for the frameworks. As well as Code Qwen and the OpenAI API as our Large Language Models, and hosted everything on the Google Cloud Platform.