Incident Analyst

Inspiration

Existing dashboards and chatbots used for debugging.

What it does

AI chatbot which connects to vector databases and enable greater insights using RAG, specifically for backend Kubernetes infrastructure troubleshooting. During troubleshooting, agents with different specialties will coordinate together to develop the best solutions.

How we built it

Incident Analyst uses a FastAPI backend and a Nextjs frontend, both deployed to Render.com. The FastAPI backend accesses and interfaces with K8s, enabling an AI-powered chatbot for Kubernetes infrastructure log analysis and troubleshooting. The AI chat feature connects to you.com to query and get information.

Challenges we ran into

The scope of this problem was much larger than expected. The existing demos were only able to capture the intention and basic user interface.

Accomplishments that we're proud of

Our team created great and impressive demo on such a short time. We are proud of our team culture and collaborative environment.

What we learned

So much about AI and teamwork!

What's next for Incident Analyst

Cleaning up and implementing a full backend. There is a lot of work needed to properly connect to kubernetes and use a vector database for managing logs

Built With

• fastapi
• gemini
• nextjs
• you.com