DevOps Engineer

📋 Description

• • Own the reliability and scalability of the AI search platform that Fortune 500 brands rely on to understand how ChatGPT, Perplexity, Gemini and other LLMs talk about their products. Every millisecond of latency you shave off and every 9 you add to uptime directly impacts how quickly our customers can react to AI-driven market shifts.
• • Architect and maintain multi-cloud infrastructure (AWS primary, Azure & GCP secondary) that ingests terabytes of daily model-output data, runs real-time NLP pipelines, and serves sub-second insights to dashboards used by marketing teams around the globe.
• • Build end-to-end CI/CD pipelines using GitHub Actions and Argo CD that take code from pull-request to production in under 15 minutes while enforcing security gates, automated integration tests, and blue-green deployments that customers never notice.
• • Codify every piece of infrastructure with Terraform and Ansible so that environments—from dev sandboxes to production clusters—are reproducible, version-controlled, and can be spun up or torn down with a single command.
• • Containerise micro-services with Docker and orchestrate them on Kubernetes (EKS), implementing HPA, VPA, and cluster-autoscaling policies that keep costs low during quiet periods and guarantee capacity during viral traffic spikes.
• • Instrument everything: Prometheus, Grafana, Loki, and OpenTelemetry traces give you real-time SLO dashboards, while alerting rules in Alertmanager page you before a customer ever feels pain.
• • Partner with ML engineers to optimise GPU node groups, schedule model-training jobs, and implement checkpoint/restore strategies that cut training time by 30 % without ballooning cloud spend.
• • Champion security by integrating OPA, Vault, and IAM least-privilege policies into every layer—container images are scanned on build, secrets never hit git, and compliance reports generate themselves.
• • Run game-day chaos exercises twice per quarter to validate failover, backup, and disaster-recovery procedures; document runbooks so any on-call engineer can bring the system back in <15 min.
• • Continuously reduce toil: if you do it twice, you automate it. Your bash/python tooling turns repetitive tasks into one-click scripts that free the team for higher-value work.
• • Mentor junior engineers through pair-programming and brown-bag sessions, raising the DevOps bar across the entire engineering org.
• • Stay on the bleeding edge—evaluate new CNCF projects, pilot serverless or edge-compute solutions, and present findings that influence Goodie’s technical roadmap for the next 12 months.