London-based, UK-wide
Supporting London companies and remote UK teams with practical automation builds.
About William Purnell
I build practical AI automation around real operational work.
I am a London-based software engineer helping businesses replace repeated admin, scattered data and slow handoffs with maintainable workflows, internal tools and AI-assisted systems.
Current focus
Automation that earns its place inside the business.
The work now sits at the intersection of software engineering, workflow design and practical AI implementation.
01
Workflow automation
Mapping repeated admin, handoffs, reporting loops and inbox-heavy processes before building the software that removes the drag.
02
AI integration
Adding OpenAI and AI-assisted workflows where they have a defined job: classification, extraction, retrieval, drafting, routing or decision support.
03
Internal tools
Building dashboards, portals and operational systems that give teams one reliable place to manage work.
04
Architecture audits
Helping architecture practices identify where AI and automation can support drawing, document, planning and project-information workflows.
Experience
A software engineering background, applied to operational problems.
My career started in a .NET development agency, then moved into product engineering at Instanda, an InsurTech company focused on modernising insurance operations.
The useful bit.
I can talk to business owners about process pain, then translate that into data structures, APIs, screens, automations and AI guardrails that developers can maintain and teams can actually use.
I work with modern engineering foundations including .NET, React, Next.js, Tailwind, MongoDB, Azure, Vercel, Stripe, APIs and OpenAI. The stack matters, but only after the workflow is understood.
My own AI product experiments, including Text2Quiz, also shaped how I think about practical AI: the hard part is rarely calling a model. The hard part is making the surrounding workflow reliable, useful and worth changing.
Approach
Clear recommendations before code. Staged delivery after that.
Start with the workflow, not the tool.
Keep humans in control where judgement matters.
Build small enough to prove value, solid enough to keep.
Make recommendations evidence-led and implementation-aware.
How projects usually start.
Most work begins with a consultation or audit: understand the current process, identify the highest-value opportunity, then decide whether the next step should be an integration, internal tool, AI workflow, prototype or roadmap.
Why work with me
A practical partner for teams that need momentum without guesswork.
The goal is to help you make better automation decisions before money is spent on the wrong tool, prototype or build.
Reliable Results
Committed to clear communication, staged delivery, and maintainable systems.
Proven Experience
Backed by real software engineering experience across .NET, React, Next.js, APIs, and AI.
Start with the workflow that is slowing you down.
Share the process, toolchain or AI idea you are considering and I can help identify the most practical next step.
Book a consultation