Built an AI DevOps Agent that explains outages like a human (launched today)
Hi AI builders đź‘‹
We just launched AI Incident Investigator — an AI agent that helps DevOps teams debug cloud incidents using natural language explanations instead of dashboards and logs.
It uses:
-
Claude as the base model
-
LangChain to structure prompts & memory
-
RAG to enrich LLM output with real-life infrastructure debugging patterns from AWS (ECS, ALB, CloudWatch)
Example query:
“Why did our staging app return 5xx errors at 2PM yesterday?”
🧠AI Output:
Deployment at 13:58 changed env var MAX_WORKERS
Instance ran out of memory
5xx errors started at 14:03
Suggest: rollback or increase container memory
We’re live on Product Hunt today!
đź”— https://www.producthunt.com/products/microtica-ai-agents-for-devops
Would love feedback from other AI devs here — especially if you’re working with LLMs for infra.
