AI DevOps Incident Copilot — Automated Incident Response on AWS by Shivam GuptaAI DevOps Incident Copilot — Automated Incident Response on AWS by Shivam Gupta

AI DevOps Incident Copilot — Automated Incident Response on AWS

Portfolio project
Shivam Gupta
An incident-response copilot for DevOps and SRE teams: when a service fails, it detects the problem, records it and alerts the on-call engineer automatically.
What I built
Automatic incident detection: catches HTTP 500s and service failures and captures stack traces, logs and severity
Real-time SRE dashboard (React + Recharts) for MTTR, service health and the incident queue
Structured incident logs in AWS CloudWatch for auditing
HIGH / CRITICAL alerts to on-call engineers via Amazon SNS (email + SMS)
Decoupled alarm ingestion from other services and CI/CD through Amazon SQS
Prometheus metrics with Grafana dashboards
Multi-stage Docker builds, deployed on AWS ECS Fargate behind an Application Load Balancer
Stack: React, Node.js, Docker, AWS (ECS Fargate, ALB, CloudWatch, SNS, SQS), Prometheus, Grafana, GitHub Actions

0 likes · 0 views · Oct 13, 2026

Detects production failures automatically, logs them to AWS CloudWatch, alerts on-call via SNS and tracks MTTR on a live React SRE dashboard.

Tags