Provenance
Every output traces back to its source — so you can always answer “why?”
Provenance is the chain of where things came from. When an agent produces a problem cluster, that node remembers exactly which evidence it was built from. When you write a PRD from that problem, the PRD remembers the problem. The result is a fully traceable line from raw evidence to final decision.
What you can do with it
- Hover any node to highlight its sources and everything derived from it.
- Click through from a ranked problem to the exact tickets and quotes behind it.
- Defend a decision in a review with the evidence one click away.
Answering “why are we building this?” in ten seconds
Ana, PM presenting to leadership
The situation. In a roadmap review, a VP pushes back on Ana’s top priority: “Is this really the biggest problem, or just the loudest?” In most tools she’d promise to dig up the data later.
What they do
- 1Ana clicks the problem cluster on the canvas.
- 2Its source evidence highlights instantly — 41 support tickets, 3 interview quotes, and an analytics funnel.
- 3She opens two of the quotes on screen.
The result. The room sees the evidence behind the call, live. The priority holds — not because Ana argued well, but because the decision was traceable to real data the whole time.
Lesson: traceability is what makes AI output trustworthy
- What it is
- Every node records where it came from (a “derived from” link) and what produced it. Nothing on the canvas is an unexplained black box.
- Why it matters
- The fastest way to lose trust in an AI tool is one confidently-wrong output you can’t verify. Provenance means you can always check the work — which is exactly why you can rely on it.
Don’t take a ranking on faith. Click through once or twice; the sources are always one hop away.
No black boxes
Because provenance is built in, you never have to take the AI’s word for it. The reasoning and the sources are always there to inspect.