Leeds, England, United Kingdom
I always love a challenge and learning new things. Outside of work I spend a lot of time playing guitar (mainly fingerstyle and J-rock) as well as drums. I regularly attend tech-focused MeetUps to improve and broaden my knowledge and get exposure to different ideas, solutions and perspectives. I also have an interest in developing my cloud platform experience to become a well-rounded DevOps engineer.
An existing service was unmaintained and undocumented and needed to be brought up to engineering standards and become a supportable service. I became the SME and was responsible for owning the Epic and ensuring that knowledge was shared. I setup a Concourse pipeline that linked to an Octopus Deploy project which enabled automated deployments for the service. To get it deployment ready, I added logging which sent logs to Kibana. OTel was configured and dashboards were created in Grafana. The service is deployed to an internal network, and is made accessible via an Akamai route exposing it to the public internet. I created a new Service Configuration Item and liaised with senior change management to communicate the change. We used AB testing to ensure the new service functioned the same as the existing one. I was appointed as the SME for a P1 incident that required me to spike an investigation into the underlying cause and propose a solution design. I presented the solution to the tech lead who approved the approach and I implemented the proposed changes to the frontend and backend.
The frontend application needed updates to comply with the European Accessibility Act. I made changes to a variety of user interface components such as screen reader focus changes, date picker enhancements and customer journey improvements using React and CSS. To monitor the impact of the code changes, we wrote them as experiments so we could compare the customer experience before and after the change. I tested the changes using VoiceOver and Evinced Web Flow Analyzer. Sky needed a new journey using existing shared components to provide new functionality to users. I made a new route and landing page for the new journey using React and CSS. For testing these additions, I used Percy for screenshot comparisons alongside Playwright, React Testing Library and Jest. I created new GraphQL mutations using TypeScript to enable this new journey, and wrote blackbox tests as well as sinon and chai unit tests. To help with integration testing I added personas to our stub server to allow us to have good test coverage of the new journey. To further my GraphQL knowledge I joined the GraphQL guild in Sky where I participate in mob review sessions on PRs as a way to share knowledge and maintain code quality and consistency.
Sky had a need for a consistent and reliable solution for generating styled QR codes to be used by anyone in the organisation. I was part of a team of three developers who designed and built a backend API using NodeJS and Koa to create CRUD endpoints for managing QR codes. We built a frontend using NodeJS and React to interact with the API and choose the QR code styling. Any created QR code images were stored in Google Cloud Storage, with metadata stored in a Cloud SQL MySQL database. We wrote infrastructure as code using Terraform. CI/CD pipelines were setup in CircleCI for both the frontend and backend repos which built, tested and linted the code, before pushing it up to Google Container Registry. We configured the Helm Charts to deploy both apps to a Kubernetes cluster via Octopus Deploy.
I was responsible for a piece of work to extract, uplift and merge data stored in Elasticsearch to improve search efficiency and support additional functionality. In my own time, as a way to develop my knowledge of SRE/DevOps, I produced a steel thread CI/CD pipeline for a Node web app using GitHub and GitHub Actions, building and pushing Docker images to Docker Hub. Using Docker Compose I stood up a local runtime environment consisting of Watchtower for monitoring for container image updates and updating the app, cAdvisor for monitoring the containers, Prometheus for scraping container metrics from cAdvisor and Grafana for visualising the metrics. CI/CD status notifications and alerts setup in Grafana were sent to a Discord channel via a webhook.