🕵️‍♂️ Katana

A hackathon project built alongside Niall, Cathal and Ivan in 30 hours to help monitor and automatically remediate issues in Kubernetes clusters using AI agents.

Katana deploys an AI agent alongside each Kubernetes cluster that pulls logs and performance metrics from running pods. This allows it to detect and remediate highly stressed servers while gaining insight into root causes through log analysis.

The project was built using a modern tech stack including Django for the frontend/backend, along with minikube, kubectl, Python, Prometheus, Docker and Flask. We focused heavily on following best practices and writing clean, scalable code.

My main contributions were finalizing the UI design of the monitoring dashboard and helping develop the Python-based AI agent that interfaces with the Kubernetes clusters.

Coming into this with no prior experience with Kubernetes or Django, I was able to learn a lot about the technologies and their applications.