markiskorova/ai-ops-assistant
π οΈ AI Ops Assistant β A Go-based backend system for automated log summarization, ticket triage, and changelog generation. Built with GraphQL, JWT auth, PostgreSQL, and a worker queue architecture for scalable operations.
π§ AI Ops Assistant
AI Ops Assistant is a cloud-native backend project simulating an AI-powered operational triage and summarization platform. Designed with scalability and team productivity in mind, it showcases engineering patterns applicable to internal tooling, data processing, and privacy-conscious automation.
Built using Go, GraphQL, Docker, and Terraform with secure JWT auth, this system processes logs and tickets asynchronously, summarizes them via OpenAI, and delivers insights via a robust APIβmaking it ideal for platforms focused on data control, governance, and observability.
βοΈ Tech Stack
- Backend: Go 1.23, GORM (PostgreSQL), GraphQL (graphql-go)
- AI Integration: OpenAI API for summarization
- Auth: JWT
- Infra: Docker, Docker Compose, Terraform (AWS)
- DevOps: GitHub Actions (CI/CD ready)
- Observability: Prometheus (metrics), Grafana (dashboards), Alertmanager (alerts)
β Key Features
- π§Ύ Summarizes logs using OpenAI's GPT API.
- π·οΈ Classifies tickets with pluggable business logic.
- π GraphQL API for querying logs, tickets, and changelogs.
- π JWT-secured authentication and modular user management.
- π§΅ Cleanly separated microservice-style API and worker processes.
- π Observability with Prometheus & Grafana:
- API & Worker metrics exported at
/metrics - Dashboards for API throughput, latency (P95/P99), error rates
- Worker throughput, failures, and queue depth
- Alert rules for error rates, latency, and backlogs
- API & Worker metrics exported at
- ποΈ Infrastructure-as-Code via Terraform (AWS RDS, ECS, IAM).
- π³ Local development via Docker Compose with minimal setup.
π System Design
See the System Design One-Pager for an architecture breakdown.
π§ͺ Run Locally
Core stack
docker-compose up --buildGraphQL API available at:
http://localhost:8080/graphql
With observability stack
docker-compose -f docker-compose.yml -f docker-compose.obsv.yml up -d --build- Prometheus: http://localhost:9090
- Grafana: http://localhost:3000 (admin / admin)
Provisioned dashboard: AI Ops β API & Workers
(Shows API & worker throughput, latency, errors, and queue depth.)
π£οΈ Project Roadmap
See the Project Plan & Roadmap for phased implementation and future plans.
Observability is part of Phase 3 β Infrastructure & Observability.
π§βπ» Author Notes
This project was created to demonstrate backend leadership and infrastructure fluency aligned with real-world SaaS tooling. Its design prioritizes modularity, secure data handling, observability, and developer productivity.
MIT License
