01
Full surface ownership.
Frontend, backend, AI systems, integrations, and design intent. Connected, not siloed. One person, one coherent vision.
Software studio, Edenvale, South Africa
Full-stack products, AI-native systems, and protocol tooling, shaped with architectural depth, visual conviction, and the kind of care that makes complex things feel inevitable.
Clippy is active. Ask about project depth, stack details, rates, or the short version.
17+
Live APIs orchestrated inside one platform
30+
AI tools unified into a single product surface
4+
Published packages, wrappers, and MCP servers
1,027
Poems archived into a searchable, discussable PWA
Why this studio
01
Frontend, backend, AI systems, integrations, and design intent. Connected, not siloed. One person, one coherent vision.
02
AI accelerates the workflow. Judgement stays human. Velocity is for going deeper, not cutting corners.
03
Ambitious products still need elegance. Flow, hierarchy, language, and the feel of a system in the hand. These are engineering decisions.
04
Protocol design, unusual integrations, unconventional product ideas, architecture from nothing. The difficult edge is where the best work lives.
Selected work
Each project demanded a different shape of thinking. The portfolio reads as range because the work required it, not because diversity was the goal.
01 / Case study
End-to-end platform architecture: frontend systems, backend infrastructure, custom API design, and a complete MCP server engineered from the ground up.
The kind of build that reveals range: product instinct, systems thinking, implementation depth, and protocol fluency in a single engagement.
Visit project
02 / Case study
A living Earth2 intelligence layer with real-time market telemetry, interactive cartography, and an AI companion hardened against platform complexity.
Seventeen APIs integrated without losing clarity. Complex integrations are where the work gets interesting.
Visit project
03 / Case study
Thirty-plus AI instruments for analysis, conversion, and experimentation, unified into one product surface that stays fast and navigable.
The challenge was not the tools themselves. It was making a broad feature surface feel coherent and effortless.
Visit project
04 / Case study
A native Windows code editor built in C and Win32. AI-assisted, Git-integrated, plugin-ready. Close to the metal by choice.
Evidence that the work here is not limited to the web stack. Hard problems. Low-level engineering. By preference.
Visit project

kinspace.co.za
A community architecture for people living with chronic and mental health challenges, built around safety, meaningful connection, and clinical resource access.
pathnote.co.za
pathnote.co.za
An audible location engine that transforms walks into narrative audio journeys using live geolocation, mapping intelligence, and AI-directed storytelling.
messagecfo.com
messagecfo.com
WhatsApp-native financial operations: invoicing, expense tracking, balance sheets, and client management through conversational interface.
platedom.com
platedom.com
AI-powered restaurant merchandising that transforms menus into high-fidelity visual experiences and stronger digital presence for food brands.
Open source & protocol engineering
Wrappers, MCP servers, and developer tooling built for real integration depth and AI-native workflows. Published, documented, production-facing.
Package / protocol
A clean TypeScript wrapper around the Earth2 API. Properties, accounts, transactions, and marketplace data made composable for downstream builds.
Package / protocol
A complete MCP server that lets AI clients access Earth2 account data, wallet activity, properties, and marketplace actions through structured tools.
Package / protocol
Built from scratch for Morphed.io. Platform APIs turned into usable AI tools with authentication, data access, and production-grade architecture.
Package / protocol
An extended HubSpot MCP implementation with deeper CRM operations, stronger integration ergonomics, and more serious backend architecture than standard demos.
Capabilities
Most developers can wire frameworks together. The difference is knowing what to build, what to leave out, and how to make complex systems feel simple.
Interfaces built with structure, conviction, and taste. Translating unclear requirements into products that feel composed from first impression to final interaction.
AI is not decoration here. It is part of the product architecture: assistants, structured workflows, model orchestration, and protocol tooling designed as first-class features.
The invisible engineering most portfolios omit: authentication, API architecture, database design, third-party integration complexity, edge cases, and the structural work that makes products hold.
Current toolkit
Equally at home shipping clean frontends, backend systems, data pipelines, and AI-enhanced workflows. The range produces coherent systems rather than disconnected pieces.
Investment
Ranges for the work I usually take on. Final pricing follows complexity, integration depth, urgency, and how much of the system falls under my ownership.
Engagement
Portfolio sites, landing pages, and compact business websites that need to look sharp and ship cleanly.
Engagement
Fuller company sites with stronger structure, more content depth, and proper lead capture.
Engagement
Online stores with product structure, payment flow, and enough polish to feel trustworthy from day one.
Engagement
Platforms, dashboards, AI-heavy products, and custom systems with deeper engineering and more moving parts.
Engagement
APIs turned into usable model tools with proper architecture, testing, and publishable packaging.
Engagement
Backend systems, auth flows, endpoint design, and platform integration work that powers the product behind the curtain.
Pricing note
When a project spans product strategy, interface design, backend architecture, AI workflows, and custom integrations, it enters the custom range. For a faster estimate, Clippy can qualify scope before you reach out.
Process
The projects change shape. The rhythm stays: find the real problem, design the structure, build with precision, refine until it reads.
Step 1
Most projects hide their actual problem behind the initial ask. Better scope and better decisions begin with finding it.
Step 2
Architecture that survives growth: clean data flow, honest abstractions, and integration points established before they become painful.
Step 3
Automation, assistants, and model capabilities added with purpose. Stronger products, not trend compliance.
Step 4
Good software communicates. Copy, motion, layout, and interaction quality are not cosmetic. They are structural credibility.
Contact
Taking on contract work, freelance builds, and teams that want a developer with real ownership range. Serious briefs. Real ambition. That is the filter.