Selected work & systems
Systems I've helped design, operate, and keep boring in production.
I work primarily as a backend / platform engineer: distributed services, AWS-based infrastructure, Kubernetes, and the tooling around them. Most of my impact comes from making it easier for teams to ship reliable, observable software in production.
Below is a curated view of initiatives and projects that best represent how I think about architecture, trade-offs, and ownership — especially in healthcare and other regulated environments.
Selected projects & initiatives
A cross-section of work that spans backend services, platform engineering, security, and frontend experience.
A cohesive approach to packaging, deploying, and operating services across environments, focused on repeatability, traceability, and operational clarity.
- Standardised how applications are packaged, versioned, and released across teams and environments.
- Introduced supporting services and tooling that improved observability and reliability for production workloads.
- Reduced manual deployment and handover steps by turning ad‑hoc runbooks into repeatable automation.
Security-focused improvements to the software delivery pipeline so issues surface early, with clear, automated feedback loops.
- Built automated security and quality checks that run as part of CI/CD, not as a separate afterthought.
- Embedded security practices into the pipeline itself, reducing reliance on manual review for regulated products.
- Aligned gates and reporting with product requirements so teams could move fast without weakening controls.
A maintained Kubernetes lab environment and supporting tooling to make container orchestration approachable for engineers and operations teams.
- Built and continuously maintained local Kubernetes clusters used for experimentation and validation.
- Created CLI tools in Python and Go to simplify day-to-day operational tasks for non-platform teams.
- Enabled safer experimentation by providing realistic, isolated environments that mirrored production topologies.
Practical patterns for understanding system behaviour in production, focused on fast feedback and clear signals.
- Implemented monitoring and alerting that shortened the feedback loop for production incidents.
- Improved observability for existing services, reducing guesswork when investigating behaviour in the field.
- Contributed to shared patterns for logs, metrics, and alerts so teams reasoned about systems in a consistent way.
A set of small, focused tools that use LLMs to assist with incident context gathering, code navigation, and runbook-style suggestions.
- Experimented with function calling, embeddings, and RAG workflows to surface the right operational context for a given issue.
- Validated where AI assistance adds real leverage for engineers and where traditional tooling remains a better fit.
- Built a foundation to bring AI safely into existing engineering workflows without over‑promising or hiding complexity.
This site — a focused, fast portfolio that reflects how I think about software: opinionated, practical, and quietly ambitious.
- Designed an interface that surfaces work history, systems, and technical depth without unnecessary noise.
- Built on a modern React/Next.js stack with attention to performance, accessibility, and simple deployment.
- Treats the site as an evolving lab for trying ideas in developer experience, content, and presentation.
Collaborated with a French BLE location provider to bring real-time indoor location into an Android app and web dashboards, then wired that data into map-based workflows so teams could act on location intelligence instead of static reports.
- Built real-time locationing into an Android app using BLE and defined the workflows that feed location data into downstream systems.
- Integrated location data and maps into web dashboards so operators could see live position and context instead of post-hoc reports.
- Connected Android locationing, BLE infrastructure, and dashboards into a single flow so location intelligence drove decisions (routing, occupancy, alerts).
Used LiveKit to enable real-time WebRTC communication between smart devices, mobile and web apps, and legacy SIP phones, including voice, video, and screensharing, so one platform could serve modern and legacy endpoints.
- Enabled real-time voice and video between browsers, native apps, and smart devices using LiveKit's WebRTC stack so users could join from any supported client.
- Bridged legacy SIP phones and displays into the same LiveKit rooms so existing hardware could participate in the same calls and sessions as web and app users.
- Added screensharing and media flows so participants could share screens and collaborate in real time alongside voice and video.
GitHub & personal projects
Outside of work, I keep a small set of experiments, utilities, and prototypes on GitHub — places where I try ideas in tooling, delivery workflows, and developer experience before they make their way into production systems.
I keep most of my day-to-day work in private repositories. Public projects and experiments will appear here as they become available. In the meantime, you can browse my GitHub profile directly.