I run production Kubernetes platforms on AWS, and I use AI to get DevOps work done faster.

Most of this engagement is Kubernetes and DevOps on AWS. I look after a dual-cluster EKS platform (UAT and production) for a fintech group: node and workload scaling, CI/CD, the observability stack, and production incident response.

• Upgraded Kubernetes clusters on premises and on EKS, checking node groups and workload compatibility before each version cutover.
• Hardened encryption in transit and at rest to a SOC 2-aligned standard, and tightened identity and network exposure as part of that work.
• Moved a queue-driven worker fleet from StatefulSets to Deployments and wired KEDA autoscaling to queue depth, so the fleet follows real demand instead of a fixed replica count.
• Fixed a CI image-tagging bug that was silently shipping stale images. Deploys now pin to the commit SHA.
• Run Prometheus, Grafana, Loki and Alertmanager, and cleaned up alert noise so pages point at real problems.
• Led recovery on a multi-day production outage, then the follow-up hardening, and wrote a blue-green deployment feasibility report.
• Resolved 30+ incidents and delivered 10+ pipeline optimisations in the first two months. Wrote 7 runbooks in the first 30 days.
• Use internal AI-assisted ops automation I built for clients to draft incident notes, update tickets and keep timesheets current. I review every output before it goes out.

Systems & technologies: AWS EKS, Kubernetes, Docker, Helm, KEDA, Bitbucket Pipelines, Prometheus, Grafana, Loki, Alertmanager, RabbitMQ / Amazon MQ, RDS PostgreSQL, S3, Mosquitto MQTT, .NET / IIS services, Python, Bash, Jira.

Embedded senior engineer on client platforms in banking, HR and legal tech, healthcare and media. Alongside feature delivery I do AWS performance and cost work, release governance, and production support inside regulated delivery processes.

• Tuned Elastic Beanstalk and nginx on a production system: worker and connection limits, timeouts and static handling, aimed at slow pages and unstable releases.
• Built a Lambda check that watches certificate expiry and raises a ticket before a certificate lapses.
• AWS performance and cost work on a large HR and legal-tech production system, and cost reviews for a tier-1 bank’s learning platform, written up inside that bank’s change process.
• Weekly releases to UAT and production across concurrent projects, with CI/CD in the path.
• Prometheus, Grafana and CloudWatch on the stacks that already used them, so production support had something concrete to look at.
• Rescued a deadline-critical media analytics project and delivered it on time.
• The same internal AI-assisted ops automation I built for clients drafts timesheets, tickets and incident triage notes. I check each draft before it is sent.

Systems & technologies: AWS (Elastic Beanstalk, Amplify, RDS, Lambda, DynamoDB, ECS, EC2, S3, CloudWatch, AppSync, API Gateway, Cost Explorer), nginx, Django / Python, React / React Native, Node.js, GraphQL, Flowable.

Joined a new real-estate SaaS build with no infrastructure in place. The DevOps part of the role was setting up the AWS footprint and the delivery pipeline alongside the product.

• Set up the initial AWS footprint: S3 storage, RDS PostgreSQL, and Lambda for background work.
• Built the GitHub-to-AWS CI/CD pipeline, with tests and SonarQube quality gates on every merge.
• Handled weekly production deployments, daily code reviews and delivery habits aligned with DORA.
• Kept building the product in the same engagement: React and TypeScript on tRPC, with Prisma against PostgreSQL.

Systems & technologies: AWS (S3, RDS, Lambda), GitHub Actions CI/CD, SonarQube, PostgreSQL, Prisma, tRPC, React / TypeScript, Jest, Storybook.