Professional delivery cases explain my responsibilities and results. Five independent demonstrators let you inspect working applications, source code and architecture decisions using synthetic data.
Public code on GitHubAnonymised customer work and a document-retrieval proof of concept. Open a case for the problem, my contribution, design decisions and outcome.
An enterprise cloud opportunity for a customer whose disconnected inventory and operational systems limited visibility and created manual work.
ResultThe approved proposal progressed into production, improving operational visibility, reducing manual processes and supporting lower operating costs.
Read the delivery exampleArchitecture and delivery of an LLM-powered platform for document analysis, extraction, contextual summaries and structured workflow responses.
ResultDeployed for multiple operational users and recurring document workloads, with ongoing development and support.
Read the delivery exampleArchitecture, operational reliability and technical support for a production vehicle-inspection and regulatory-information platform.
ResultImproved stability, reduced recurring incidents and made support more efficient through database query optimisation, monitoring, exception handling and consistent integration patterns.
Read the delivery exampleA data-integration and quality-management solution for pharmaceutical manufacturing, supporting Continued Process Verification and quality investigations.
ResultImproved manufacturing-data visibility, standardised quality reporting and reduced manual consolidation.
Read the delivery exampleA working document-based RAG application with local model inference, source visibility and retrieval metadata for manual verification.
ResultA working proof of concept for testing document retrieval and checking generated answers against their sources.
Read the delivery exampleEach has its own live application, repository, usage guide and architecture walkthrough. These synthetic demonstrations are separate from customer delivery.
Explore a fictional knowledge base, inspect citations and compare retrieval decisions. Optional local Ollama inference adds a separately validated model draft.
Explore project & architectureSend signed synthetic events through real HTTP, inject receiver failures and recover with visible retries, deduplication and dead-letter history.
Explore project & architectureExplore a fictional customer brief, compare feasible architecture options and produce an editable proposal with explicit assumptions, costs and a PoC plan.
Explore project & architectureIngest synthetic manufacturing measurements, investigate known quality defects and follow a reviewed correction from its original source to its current overlay.
Explore project & architectureInvestigate synthetic orders through real MCP tools, inspect the proposed follow-up and approve one bounded, replay-safe task write.
Explore project & architectureExplore Math For Maya, my maths practice app for children, with its original app guide and support pages.
A colourful maths practice app that helps children work through addition and subtraction, one column at a time. Clear prompts, hints and optional celebrations keep practice relaxed.
