Build AI-powered products with your product and tech teams

Every product team has an AI prototype that impressed the room and never shipped. The reasons are usually the same: the data it needs lives in six places and none of them are trustworthy, and the engineering process that builds the rest of the product does not know what to do with a model.

Building AI products is mostly not about models. It is about data you can reach and trust, and an engineering practice that treats AI as a normal part of shipping.

• Run the data modernisation track. A data platform with clear ownership, quality checks and lineage, built so product teams can use it without asking for permission.
• Grow the data ecosystem around it. Well-defined data ownership across the organisation, plus the contracts, conventions and shared datasets that let teams build on each other's work instead of rebuilding it.
• Set up the shared tooling stack. The MLOps and AI tooling every team uses, so that a new AI feature starts from what exists rather than from scratch.
• Put AI in the developer workflow. Coding assistants and agents introduced team by team, with the practices that make them a net gain: review, testing and clear boundaries.
• Keep an eye on what is coming. The most valuable techniques are rarely something you can buy from day one. I set up the teams to test new techniques against real problems and to carry the ones that hold up into the roadmap.
• Get the first AI products out. Pick the use cases with a real path to production, and steer them from prototype to something with monitoring and an owner.
• Coach the leads. Product managers and tech leads who can scope, evaluate and ship AI features on their own.

I ran a data modernisation track that built a new data platform and the data ecosystem around it. That foundation is what let product and tech start building AI products. In parallel, I introduced agents and coding assistants into the developer workflow at team level, changing how teams build rather than just adding a tool.

• A data platform product teams build on, with owners and quality checks.
• Engineering teams that use agents and coding assistants as part of their normal workflow.
• AI features in production, with monitoring and someone accountable for them.
• Product and tech leads who can take the next AI product from idea to release without me.