The work is
the argument.
Every case study is a real engagement. Real problem, real organisation, real outcome. Specificity is the only form of credibility that cannot be manufactured.
Seamfix UK
NHS Workforce Access Platform
Building an NHS-grade workforce access management platform for care organisations across the UK.
View Case StudyShopify
Checkout Fee Application
A three-layer Shopify application for configurable checkout fees with built-in buyer agreement.
View Case StudyBeticlaw
Conversational AI Sports Agent
Architecting a conversational AI agent for sports betting intelligence — monorepo, mobile, and backend from the ground up.
View Case StudyConfidential
GEO Strategy Playbook
A comprehensive GEO playbook for AI-era search visibility — covering WooCommerce, structured content, and LLM discoverability.
View Case StudySeamfix UK
NHS Workforce Access Platform
Building an NHS-grade workforce access management platform for care organisations across the UK.
Seamfix UK provides workforce access management solutions to NHS bodies and care organisations. Their product needed a front-end platform that could meet the accessibility, regulatory, and design standards demanded by UK healthcare — without the cost and timeline of a large enterprise agency.
The challenge was not just technical. NHS-facing platforms must be accessible to WCAG standards, intuitive for clinical staff who are not technology-first users, and deployable in a managed hosting environment with specific staging and production requirements.
Architected and built a custom WordPress theme designed from the ground up for the NHS and care sector context. This was not a template adaptation — it was a purpose-built platform with a full design system, component library, and accessibility-first implementation.
The theme was deployed to WP Engine in a staged pipeline, with a structured handover enabling the Seamfix team to manage and extend the platform independently.
THE OUTCOME
- NHS-compliant design system delivered and documented
- WCAG accessibility standards met throughout
- Staging-to-production deployment pipeline established
- Platform handed over with full documentation
- Client team able to operate independently post-engagement
STACK
Shopify
Checkout Fee Application
A three-layer Shopify application for configurable checkout fees with built-in buyer agreement.
The client needed a Shopify application that could add configurable extra fees at checkout, require buyers to agree to specific terms before completing their purchase, and be managed entirely by merchants through an admin panel — without developer involvement after deployment.
The Shopify ecosystem presents specific architectural constraints: fee logic must operate at the cart transformation layer, buyer-facing UI must operate within Shopify's checkout extension framework, and merchant configuration must persist through Shopify's data model.
Designed and built a three-layer architecture from scratch:
Layer 1 — Cart Transform Function (Rust): Performance-critical fee calculation logic built in Rust, running at the cart transformation layer.
Layer 2 — Checkout UI Extension (React): A React-based checkout extension providing the buyer-facing fee display and agreement checkbox.
Layer 3 — Admin Panel (Remix + Polaris): A merchant administration panel enabling fee configuration without technical knowledge.
THE OUTCOME
- Three-layer architecture fully integrated
- Merchant configuration without developer involvement
- Fee logic performant at cart transformation layer
- Full documentation for merchant onboarding
- Application published and live
STACK
Beticlaw
Conversational AI Sports Agent
Architecting a conversational AI agent for sports betting intelligence — monorepo, mobile, and backend from the ground up.
Sports betting is a high-frequency, information-dense domain where the quality of a decision depends on synthesising multiple data streams in real time. Most bettors operate on instinct, partial information, or opaque third-party tips.
The opportunity was to build an AI agent that could reason across sports data, surface actionable intelligence, and engage users through natural conversation — on mobile, natively.
Backend — Modular Python Monolith: A Python backend structured across five service layers: conversation management, data ingestion, reasoning/agent orchestration, user management, and API gateway.
Mobile Application — React Native / Expo: A cross-platform mobile application built with React Native and Expo, providing the primary conversational interface.
Infrastructure — pnpm + Turborepo Monorepo: All applications managed within a single monorepo, enabling shared packages and consistent tooling.
THE OUTCOME
- Modular monolith architecture for future scaling
- Cross-platform mobile application in testing
- Agentic reasoning pipeline in development
- Monorepo infrastructure enabling rapid iteration
STACK
STATUS
Active development
Confidential
GEO Strategy Playbook
A comprehensive GEO playbook for AI-era search visibility — covering WooCommerce, structured content, and LLM discoverability.
Search behaviour is changing faster than most marketing teams can track. AI-powered search engines — Google SGE, Bing Copilot, Perplexity, ChatGPT search — are increasingly intermediating between content and the people who need it.
For e-commerce businesses in particular, the risk is acute: if your products are not structured in ways that AI systems can parse, understand, and cite, you lose organic visibility.
A comprehensive GEO audit and implementation playbook covering:
→ AI search visibility assessment for existing content → WooCommerce SKU and product data optimisation → Schema markup strategy for AI-era structured data → Content architecture for Generative Engine discoverability → Implementation checklist prioritised by impact
THE OUTCOME
- Full GEO playbook delivered as publication-ready document
- Implementation checklist immediately actionable
- AI-era search framework applicable across product catalogue
- Client team upskilled on GEO principles
STACK
Not all work is
published here.
Engagements involving sensitive commercial contexts, ongoing products, or clients who have requested confidentiality are not listed publicly. If you would like to know whether we have relevant experience, ask us directly.
Ask About Relevant Experience