It’s kind of meta when you think about it. You can then deploy these models to Databricks Model Serving which provides the actual ‘brains’ of the system you build (without you having to maintain any of the model serving infrastructure). The Databricks Extension enables all the code you develop on VS Code to seamlessly run on Databricks compute clusters (or it could be a single machine with plenty of memory, if that’s what’s needed) and facilitates logging of models with MLFlow.

In this blog, I will walk through how you can build your own AI coding assistant BrickPilot (😁) with Visual Studio, GitHub Copilot and the Databricks extension for Visual Studio Code. Also, you can still leverage Copilot to enrich your coding process.

This ensures the development process and the model serving are scalable and governed. With this, a developer can develop entirely in VS Code while using Databricks compute and deploy models to Databricks. You can fill this gap with the Databricks extension for VS Code. However, to build and deploy AI based systems such as GitHub Copilot, you need a lot more than a code editor and an intelligent coding assistant. Particularly, paired with all the nice features developers love about Visual Studio Code, Copilot offers a compelling developer experience. Tools built with large language models are revolutionizing the way we interact with systems and GitHub Copilot, powered by OpenAI’s Codex, is a great example. Build your own AI Coding Assistant with Databricks using Visual Studio Code and GitHub Copilot

Visual studio code copilot