Building useful software for messy real-world work

I work from unclear goal to running service: framing the problem, choosing the architecture, building the thing, and keeping it understandable for the team that owns it next.

Explore the work See university platforms

Anthony Humphreys

Senior developer

interface Founder extends SeniorDev {
  range: 'idea to shipped system';
  edge: 'judgement + delivery';
}

const anthony: Founder = {
  builds: ['AI', 'mobile', 'cloud'],
  leads: 'discovery to delivery',
  brings: 'clarity to vague work',
  bias: 'make the useful thing real'
};

Range

AI, mobile, web, cloud

Mode

Unclear brief to running service

Useful softwareAI systemsMobile platformsResearch toolsService designAWS delivery

Capabilities with receipts

The work spans student services, AI assistants, public research sites, internal tools, subscription apps, and the unglamorous engineering that keeps them usable after launch.

Student-facing mobile platforms

Modernising iLancaster with React Native, Expo, faster check-in flows, safety features, digital passes, and data-informed iteration.

Applied AI that has to behave

Building tools like LUCA that support real student workflows: CV review, interview practice, job description analysis, and progress tracking.

Research and innovation delivery

Turning academic and partner ideas into usable software: maps, public platforms, data tools, admin systems, and prototypes.

Architecture that can be owned

Taking ideas from discovery and workshops through implementation, deployment, feedback, and maintenance without losing the people who run it.

Architecture and delivery

I like the bit where vague goals have to become usable systems, sensible architecture, and delivery plans that a real team can survive without forming a support group.

AWS Certified Solutions Architect - Professional gives me a structured way to reason about reliability, security, cost, scalability, migration, and operational support.

Problem framing

Turning uncertain ideas into practical options, including what to test first and what to leave alone for now.

AWS architecture

Designing secure, cost-aware, maintainable cloud systems with reliability, migration, permissions, and operational support in view.

Delivery judgement

Keeping implementation, observability, release confidence, handover, and day-two ownership connected from the start.

Applied AI systems

Building AI features around usefulness, trust, evaluation, privacy, and cost instead of letting the demo write cheques the service cannot cash.

Builds and experiments

Lexio is where I ship focused software: small apps, AI workflows, developer tools, and prototypes that prove or disprove an idea in the open.

Software studio

Lexio

A home for subscription apps and bespoke builds for developers, creators, internal tools, prototypes, and integrations.

Keeping small builds tied to a real user job, a testable assumption, and enough restraint that the experiment can actually ship.

AI tool

GMprentice

A sandbox for tabletop GMs to practise with AI adventurers: party generation, character voice, dice, initiative, secrets, and session flow.

Focused on rehearsal and feedback for live facilitation, not another content generator wearing a wizard hat.

Personal workflow tool

JobMatch AI

An AI-powered job board that parses a CV, learns preferences, searches for relevant roles, and improves from user feedback.

Testing whether matching, explanations, and feedback loops can make job search less like feeding a CV into a paper shredder.

University platforms and research translation

A lot of the interesting work happens between disciplines: careers, mobile services, sustainability, physics, public science, research visibility, and AI that has to be useful after the demo.

Careers AI assistant

LUCA

A Lancaster University careers pilot for independent employability practice: CV and cover letter review, interview modes, feedback reports, and LinkedIn recommendations.

Applying AI in a university service context, balancing usefulness, trust, safety, evaluation, privacy, and operational cost.

Mobile platform

iLancaster

A daily companion for campus life, evolved through co-production and data: Expo migration, performance work, check-in improvements, digital passes, safety tooling, and support visibility.

Maintaining and evolving a high-visibility student platform where reliability, accessibility, release confidence, and institutional service ownership matter as much as feature delivery.

Research interface

IceHunter front-end

A map interface for iceberg detection research, turning satellite radar and AI outputs into accessible location data for potential maritime use.

Translating specialist research outputs into a usable interface where uncertainty, map interaction, and public-facing interpretation need careful handling.

Research tool

Sustainable packaging tool

A web tool helping eCommerce businesses make data-driven packaging decisions that balance cost, compliance, and environmental impact.

Turning research and compliance complexity into decision support that helps businesses compare trade-offs instead of drowning in inputs.

Public science platform

AuroraWatch UK refresh

A modernisation project for a public-facing aurora platform built around real-time and historical data, alerts, and wider access to science.

Modernising a public science service around access, data clarity, reliability, and a wider audience than the people already fluent in the domain.

Research web platform

Prob_AI research hub

A public site for an EPSRC-funded AI research hub, built with Next.js, Mantine UI, AWS, university content integrations, and a lightweight admin workflow.

Making collaboration, content ownership, AWS hosting, and long-term maintenance practical for a distributed programme.

How I tend to work

Strong technical work is rarely just code. It is judgement, collaboration, trade-offs, communication, and enough taste to know when an abstraction has wandered into self-importance.

Start with the shape of the problem before picking the stack.
Design for the people who will actually use and maintain the thing.
Use AI where it expands capability, not where it decorates a demo.
Keep systems understandable enough that future teams can safely change them.
Prefer evidence from users, data, and delivery over theatre. Theatre has excellent lighting and terrible uptime.