Nova CI

Inspiration

After working on numerous projects, we started thinking about what we could build for a more developer-forward audience. One problem kept coming up: broken CI pipelines. A small mistake in a .gitlab-ci.yml file can break a pipeline, and developers usually only find out after it fails. Then they have to dig through long logs just to understand what went wrong. That frustration is what inspired Nova CI.

What it does

Nova CI is an AI-powered CI assistant with a chat interface. Developers can ask questions about their CI pipelines, understand why a pipeline failed, and get suggestions for fixing configuration issues. It analyzes CI files and pipeline logs to provide quick explanations and guidance.

How we built it

Nova CI is built using Python and FastAPI. It integrates with GitLab to read CI configuration files and pipeline logs. We use YAML validation for structural checks and an LLM to analyze pipeline logic and failures. A chat interface sits on top, allowing developers to interact with the system conversationally.

Challenges we ran into

One of the biggest challenges was interpreting CI pipeline logs, which can be long and inconsistent. Another challenge was making sure the AI responses were clear and actually helpful for developers.

Accomplishments that we’re proud of

We built a working system that can analyze CI pipelines and explain failures in plain language. We’re also proud of the chat interface, which makes interacting with CI analysis much simpler.

What we learned

We learned a lot about CI/CD pipelines, GitLab integrations, and applying AI to developer tools. We also saw how useful conversational interfaces can be for technical debugging.

What’s next for Nova CI

Next, we want to improve the accuracy of the analysis, support more CI platforms, and eventually allow Nova CI to suggest or generate fixes automatically perhaps through the command line alone.