Frontend engineer with a background in test, building accessible interfaces and AI-enhanced products. React, Next.js and Node: going deeper, not wider.
I started on an internship at CloudM, adding an accordion menu to a knowledge base. Small feature, but I got completely absorbed in how the logic fitted together and how it ought to behave, and that settled it. I’ve always liked problems and puzzles, and this is just the version I never got tired of.
A good stretch of my career since has been as a Software Engineer in Test (front-end at Travelchapter, back-end at JLR), and it left me with a conviction I haven’t shaken: the best user journey is usually the simplest one. Nobody wants to click through five things to get somewhere. It turns out the qualities that make a journey pleasant to use are the same ones that make it possible to test: honest semantic elements, predictable states, hooks where you need them. Building for testability and building for people pull in the same direction far more often than people assume.
So I care a great deal about accessibility, and about genuinely meeting the W3C and a11y standards rather than gesturing at them. Get that right while still building something that grips people, and you’ll have very happy users.
I’m not trying to be a jack of all trades. I’d rather master a core stack (React, Next.js, Node and Express) and extend outwards deliberately. Back-end was the part that frustrated me most, so I went at it directly; being able to contribute across a product instead of one layer of it changes the kind of work you can take on. Cloud, with AWS and GCP, is where that’s heading next. A lot of that deliberateness comes from reading. The Pragmatic Programmer has probably shaped how I think about building software more than any single project has.
Most recently that’s meant AI. I was part of a team that built Ask Andi, a chatbot live on AND Digital’s site that matches prospective clients to the right expert, which meant prompt hardening, abuse guardrails, and a great deal of work coaxing an LLM into responses that reliably conform to a JSON schema. That last part was harder than I expected and improved the product enormously. I’ve been reading around it since: Architecting Agentic Solutions with Claude most recently, which fed into sitting and passing Anthropic’s Claude Architect Foundations exam. AI-enhanced products are where a lot of the genuinely interesting work sits right now, and it’s the work I want more of.
Built an AI-powered chatbot widget for AND Digital's website that discusses the consultancy's capabilities with prospective clients and routes them to the right expert.
Built the front-end upload flow for a platform that lets team managers distribute licensed media assets to customers.
Tested the API services behind JLR's OneApp and Home Charging Units products, covering happy and unhappy paths ahead of production releases.
Built and maintained a Cypress test suite spanning two newly merged flagship sites, plus a recurring data-integrity script that caught listings missing required property attributes before they reached customers.
AND Digital
Consultant software engineer at AND Digital, delivering full-stack development and testing work across a range of client engagements.
View roleCloudM
Intern web developer contributing to the front-end redesign of CloudM's customer knowledgebase, picking up Sass and SCSS along the way.
View roleAnonymised to role only. Full attributed references available on request.
Quote 1 of 6
Employed, not currently job hunting. Find me on LinkedIn and GitHub, or drop an email. I read everything.
Or copy it directly: jakeshaw121@gmail.com