As a cross-stack developer, I quickly build functional prototypes and AI-powered tools that help teams validate ideas early and move from concept to working software with minimal friction. My focus is on creating usable, realistic implementations that reveal technical constraints, surface design considerations, and give stakeholders and developers something tangible to test and iterate on.
I approach building as both a creative endeavour and an extension of analysis: turning clarified intent into working code, exploring feasibility, and identifying the technical decisions that matter before full-scale engineering begins. Whether I’m shaping a small proof-of-concept or assembling a more complete internal tool, I work end-to-end — from user flow to front-end behaviour, data modelling, API design, and deployment.
The core of my build practice includes:
- Rapid prototyping to validate ideas and de-risk assumptions
- AI tooling & automation, including workflow-specific LLM integrations
- Full-stack development (React, Typescript, TailwindCSS, Vite, Node, Php, Laravel, MySQL)
- Front-end UI/UX design with a focus on interaction, performance, and UX clarity
- Design systems that maintain consistency and reduce cognitive load
- RAG / LLM toolchains for retrieval, reasoning, and generation workflows
- D3 & data visualisation for interactive charts, timelines, and structural maps
- Deployment pipelines using DigitalOcean, Nginx, and GitHub Actions
- Technical feasibility analysis embedded directly in working prototypes
- Bridging analysis and implementation so specifications turn into coherent, testable software
Across multiple decades of web design, front-end/back-end engineering, design systems, and devops, I’ve worked through every major shift in the ecosystem — from hand-rolled HTML to template engines, from jQuery to component frameworks, from FTP uploads to CI/CD pipelines, and now into AI-accelerated, multimodal development.
Building with AI
I stay fully across the current “vibe-coding” landscape: command-line tools that scaffold code, generate components, refactor files, and maintain project structure; model-driven coding environments that let me iterate faster and with greater accuracy; and hybrid workflows where AI assists both the technical build and the conceptual framing.
Building conversational Interfaces
I also use AI to create interfaces that let users use AI to manipulate complex software systems — Gantt charts, task lists, flow diagrams, taxonomies — through a conversational layer that maps natural language into structured operations on objects in the UI.
I work on tools like reSpec, which combine CLI-based code generation with content authoring: the same command-line interface can update the spec content and regenerate the UI that displays it. This feels like a new mode of co-authoring — where code, content, and interface evolve together, and AI acts as both collaborator and compiler.