Agentic Engineering
Agentic coding your tech lead can trust
Your team is already using AI to write code. I help you set up the skills, review gates and tests that make that code safe to ship.
The problem
More code is not the same as more progress
Agentic coding tools can write a feature in minutes. Then someone has to review it. Pull requests get bigger, reviews get shallower, and the tests were often written by the same agent that wrote the code, so they check what it did rather than what you meant.
That's how teams end up with more code, the same review capacity, and a creeping worry that nobody fully understands the codebase any more.
Strategic programming
People own the intent. Agents write inside the lines.
Strategic programming splits the job cleanly:
People own
the architecture, the intent of each change and the tests that prove it.
Agents write
the code inside those boundaries, using skills your team has written down.
Review gates
check the change against the intent, not just for style.
The tests come first and come from a person. That one rule does more for trust in AI-written code than any tool choice.
ZPD for teams
Seniors multiplied, juniors grown
With the right scaffolding of agent skills, review gates and a test harness, each engineer can work one level above where they are today. Seniors spend their time on design and review instead of boilerplate. Juniors and mid-levels take on work they'd otherwise wait a year for, with guardrails that catch mistakes early.
You get the output of a bigger team for the cost of a smaller one, and the people you have get better faster.
What you get
The Agentic Engineering Enablement programme
- A team playbook for agentic coding: when to use agents, how to brief them and what never goes to an agent
- Agent skills written from your codebase and conventions, so the agent codes the way your team does
- Review gates for AI-written changes: size limits, intent checks and who signs off what
- A test harness for AI-written code: human-owned acceptance tests, plus the checks that stop an agent grading its own homework
- Pairing sessions with your engineers on real tickets, not toy examples
4–6 weeks. Typically £6,000–£15,000, depending on team size.
Codebase review
Start with a 30-minute codebase review
Show me how your team uses AI coding tools today, on a real repo and real pull requests. I'll tell you where the risk is and what I'd change first, and send you the test-harness checklist afterwards, whether or not we work together.
Proof from my own production lines
I build my own platforms this way
I build and maintain my own .NET and Next.js platforms with coding agents, working from written specs, handovers and skills. The pipelines behind Campercation and TrustedVillas run in production on that code every week.
Campercation
A campervan travel platform I own and run
The blog runs on a search-seeded pipeline: topics come from search data, drafts are generated and GEO-structured, images go to a CDN and posts publish through a .NET API. A weekly SEO loop decides what to write or update next.
25,000+ stopovers on the live map
RunningTrustedVillas
A villa rental platform I own and run
A property sync pipeline with monitoring, so listings stay current without anyone re-keying them. Hub content sits under change control, so an agent can’t quietly rewrite a live page.
63,000+ properties synced
RunningVerqos
The content engine behind both
The same pipeline, productised. It plans, drafts, checks and publishes, with a human review queue for anything legal, pricing or competitive.
Hundreds of hours saved
Running
FAQ
What is agentic coding?
It's software development where an AI agent does more than autocomplete. It plans a change, edits several files, runs the tests and iterates. Done well, engineers set the intent and the tests and the agent does the typing. Done badly, you get a lot of code nobody has really reviewed.
How do you test AI-written code?
Write the acceptance tests yourself, before the agent writes the code, so the tests check what you meant rather than what it did. Then add review gates: small changes, a stated intent on every pull request, and a person who owns the merge. My test-harness checklist covers the rest.
What is AI code review?
It's using AI to review code changes, or putting AI-written code through a review process designed for it. Both help, but neither is a substitute for a person who understands the intent. I set up the second first, then use the first to speed it up.
What's the difference between agentic engineering and vibe coding?
Vibe coding is prompting until something works and shipping it. Agentic engineering is the disciplined version: people own the architecture, intent and tests, and agents write code inside those limits. One is fine for a prototype. The other is how you run production software.
Do we still need junior developers if we use AI coding agents?
Yes, and they can grow faster than before. With the right scaffolding, agents let juniors take on work a level above their experience while seniors review and design. The teams that do well treat agents as a way to multiply people, not to cut them.
Which AI coding tools do you work with?
Whatever your team already uses. The playbook, skills and review gates are about how you work, not which tool you pay for, so they carry over if you switch.