AI-augmented discovery toolkit for product teams

Product discovery generates a lot of material — workshop outputs, research notes, problem statements, and design drafts. In many teams, the connection between that material and what eventually ships is weak. Insights accumulate in separate tools, and by the time a backlog is written, the original reasoning behind feature decisions is hard to trace.

Product designer Filipa Lacerda describes a system built to address this: a discovery toolkit augmented by an AI copilot, designed to keep the link between research and delivery intact across the full product development cycle.

How the toolkit works

The core of the system is a set of structured discovery artefacts — templates and formats that give workshops consistent, machine-readable outputs. With structured inputs, an AI copilot can do useful work afterward: summarizing insights, identifying recurring themes, flagging gaps in the problem definition, refining product goals, and suggesting user story candidates for epics.

The flow runs in sequence: a discovery workshop produces structured outputs, those outputs define main epics, epics scope an MVP, and the MVP breaks into user stories ready for implementation. At each step, the AI copilot helps process the material rather than generate it from scratch.

The key distinction is that discovery remains a human activity. The toolkit does not replace workshops or collaborative sense-making; it gives those sessions a format that AI can work with afterward, so the outputs do not get lost between tools.

What it does not address

The article does not cover specific tooling recommendations in depth, nor does it address how to run the underlying discovery workshops. It assumes teams already have a working discovery practice and focuses on how AI fits into the synthesis and handoff stages that follow.

Who it is useful for

PMs and product designers who work in organizations where discovery outputs consistently get lost or fail to shape what gets built. The approach applies to teams of any size, though it is most relevant where multiple people contribute to discovery and results need to be traceable across handoffs.