AI engineering and practical applications for business growth
What companies really risk with AI built software: invisible security flaws, agents with production keys, critical paths that change by accident, architecture debt, compliance, insurance exclusions and lock in. And how an experienced developer uses AI to deliver faster and for less.
Four acronyms for four outcomes: found, lifted, cited and recommended. What SEO, AEO, GEO and AIO each mean, what actually moves each one, how they stack, what the lost clicks cost in numbers, and how to measure all four without buying a dashboard.
Fixing vibe coded software has become a business of its own. Why AI built projects with no tests and no documentation break, what a rescue involves hour by hour, and why the bargain quote so often costs the most.
What the dot com crash really looked like, what the biggest AI spenders report in their own filings, where AI genuinely earns its keep, and the hidden costs a business should price in before it spends.
Where AI built prototypes break when real customers arrive: the database first, then the framework, then the language. What each popular framework is really for, the warning signs to check, and how to recover.
Two vendor integrations, one month, more than £35,000 in costs to my clients: what unreviewed AI generated code looked like in production, what it cost, and how to protect your side of an integration.
Seven in 10 carts are abandoned and a sixth of US retail sales come back. Two predictive systems, an abandonment risk scorer and a return risk predictor, built in Python and wired into Solidus, with the business case worked through properly.
How dynamic pricing actually works for a shop: estimating elasticity from your own orders, the markup rule, a demand model, the guardrails, and the UK price transparency and competition rules, with Solidus and Rails code.
Shoppers increasingly ask an assistant instead of typing keywords. What that means for a shop's product pages, descriptions and structured data, with UK dates from Google and a Rails example.
SARIMA, Prophet and gradient boosted trees compared for a shop's weekly sales: when each earns its place, how to test them honestly, and how a forecast becomes a reorder point in Rails and Solidus.
Germany and Austria share the EU AI Act and the GDPR but not the enforcers. Switzerland has neither, and a data protection act that already covers AI. What a British business needs to know, and how to build one Rails codebase for all three.
A developer's walkthrough of the processing contract, data protection by design and the impact assessment for AI features, under the UK GDPR and, for EU customers, the EU GDPR. The law first, then the Rails code.
Deleting a customer's rows is easy. Their influence on a trained model is not. What Article 17 says in the UK and the EU, what regulators have actually said about models, what machine unlearning research can do, and the Rails architecture that keeps you out of the problem.
The Digital Omnibus moved the high risk rules to December 2027 and August 2028, but the transparency duties already apply. Here is where a SaaS product sits, what it owes and how to build for it on Rails.
Most marketing automation is still a timer with a mailing list. What changes when models decide instead: guarded dynamic pricing, recommendations that read context, churn prediction that acts in time, and campaign agents, built on Rails and Solidus within UK data protection rules.
Most AI projects stall on messy data, old systems and no clear first step. The right order of work, three patterns for adding AI to a Rails and Solidus stack without replacing it, what MCP changes for legacy systems, the data protection basics and five realistic places to start.
A chatbot answers your questions, an agent does your work. What really changed between 2024 and 2026, from computer use and MCP to agent payment protocols, what agents do today, and what it means for a shop on Rails and Solidus.
What an AI agent actually is, how it differs from a chatbot and an automation rule, and how to build agents for stock, support, routing and pricing on Ruby on Rails and Solidus with the guardrails that keep them safe.
How to add product recommendations to a Solidus shop: co occurrence, collaborative filtering with ALS, content based filtering with TF IDF and a hybrid of the two, with working Python and Rails code, an A/B test and the UK consent rules, using an invented Bristol tea merchant.
What Test Driven and Behaviour Driven Development are, which Rails tools do which job, what to test before every release, and how to use AI for tests without handing it the judgement.
AI and machine learning get used as synonyms, but they are not. What each one is, how a model actually learns, what each cannot do, the Python toolkit, and why the label matters for your projects and under the EU AI Act.
What happens when rival shops run the same clever algorithm against each other? Bidding wars, the Nash equilibrium, why learning agents chase a moving target, and where optimisation turns into algorithmic collusion.
Cloud AI, self hosted models or a hybrid of both: what each can do, what goes wrong when staff paste the wrong thing, what the GDPR and the EU AI Act actually require, and how to route sensitive work so it never leaves the building.
My view as a developer on whether AI tools replace developers, set next to the controlled studies: where AI makes coding faster, where it makes experts slower, and why the responsibility stays with a person.
Performance Max and Meta's Advantage+ took the old levers away, but not the control. What still steers them, what the reports now show, and how a holdout tells you what the campaign really adds, shown on an invented Bristol coffee roaster.
ChatGPT has no ranking list, just two routes into an answer and one documented gate. What OpenAI, Microsoft and the research actually say, what is unknown, and how to measure your place.
ChatGPT search and Google's AI features fetch and cite pages by different routes, opened by different switches. Here are both routes, what the UK usage figures really say, and a plain monthly way to measure whether either tool cites you.
Google does not punish AI written pages, it punishes pages made to game the rankings. What its guidance and spam policies really say, why AI detectors fail, and how I use the tools myself.
1%of every invoice goes to a UK charity you pick.
A donation, never sponsorship. You choose the cause at onboarding.
The story behind the pledge →Questions about pricing, contracts or how we work together?
Read the FAQ
