Product engineering · OpenShift · Operations

Regulated Engineering Platform

A domain expert set the direction and priorities. I turned them into working product increments and owned the engineering path into production: React interfaces, GraphQL APIs, identity and access, data operations, and OpenShift delivery.

Owned scope

From domain need to running system

I took responsibility for the product experience, application, data operations, and delivery together.

  1. 01Product experienceWorkflows iterated with the domain expert
  2. 02Application & APITypeScript, React, GraphQL, identity, and roles
  3. 03Data & operationsJobs, storage, status, backup, and restore
  4. 04DeliveryAzure Pipelines, Argo CD, and OpenShift

Ownership visible in the system

My work covered product code, delivery, and operations.

Sustained ownership
16+ monthsTraceable product and platform work in the same system since February 2025.
TypeScript workspaces
17Frontend, API, shared libraries, and engineering tools under one build system.
Application workloads
5Application services covered by the shared OpenShift delivery process.
Deployment stages
3Separate GitOps configuration for development, pre-production, and production.

An established engineering product

The platform supports demanding domain workflows. Product interfaces, GraphQL APIs, a substantial domain model, and several operating services had grown over time.

The platform also needed a reliable OpenShift delivery process. Product development continued throughout, so application work and platform modernisation had to move at the same pace.

Working model

Domain direction, complete engineering delivery

The customer set the outcome and priority. I owned implementation across the application, delivery, and operations.

The customer brings

  • Domain knowledge and the intended outcome
  • Priority and operating constraints
  • Feedback on working software

I deliver

  • Product experience and backend behaviour
  • Migration, delivery, and operations
  • Tests, engineering controls, and maintainable code

The customer set the direction. I delivered the next production-ready increment.

The domain decisions and priorities came directly from the customer. I broke each requested change into executable steps and built it through frontend, backend, and operations.

We reviewed each working increment together. The feedback shaped the next step immediately, keeping the implementation close to the customer's real workflow.

Ownership across the stack

I translated each domain need into architecture and code, closed the gaps between system layers myself, and carried the change into a deployable state.

For OpenShift compatibility, I adapted container images and build processes, unified the workspace build system, and connected Azure Pipelines, Helm, and Argo CD in a traceable release process.

Release process

Four verifiable steps to OpenShift

The release process connects source changes, quality controls, and versioned deployment configuration.

  1. 01

    Build the change

    Azure Pipelines produces reproducible application artefacts and container images.

  2. 02

    Verify quality

    Linting, tests, and build checks stop defective revisions before delivery.

  3. 03

    Version the deployment configuration

    Helm configuration describes workloads and dependencies for each deployment stage.

  4. 04

    Deliver through GitOps

    Argo CD reconciles the approved state into OpenShift.

What reached the product

I integrated OIDC SSO and group mapping into the core product and a second product interface. Explicit error states told users when the identity configuration had failed.

Version and runtime status became visible across four components. Neo4j gained controlled backup and restore flows, an automatic safety backup before restore, an operational escape path, and integration tests for failure scenarios.

Product work continued in parallel: job management, a storage browser, traceable role sources, visible change history, and the foundation of a new tabular engineering workspace.

Outcome

The platform gained new product capabilities while we stabilised delivery and operations. The customer retains the domain decisions and priorities.

I remain responsible for an engineering task until the code is built, tested, deployable, and operable by the team. The ongoing partnership keeps knowledge and operating responsibility inside the customer organisation.

Need someone to own the engineering work?

Describe the product, current constraint, and intended outcome. I will assess which technical responsibility I can own with your domain expert through to operations.

Discuss the responsibility

Analytics settings

With your permission, I use basic analytics to see how the site is used and what I can improve. Your choice is saved for six months and can be changed at any time in the footer. Privacy policy