Find the angle each model misses.
Drop in the data, the document, or the deck. Three minds, three readings, one synthesis worth your time.
The spike correlates with the price change shipped in week 6, not onboarding. The cohort that hit month-3 just after that change is 2.4x more likely to churn.
Two factors compound: the pricing change and a 14% slower TTV in that cohort. Controlling for plan, pricing alone explains roughly 60% of the lift.
Worth checking the geography split. The non-US cohort moves the average; US churn is flat. The pricing change landed without local FX adjustment.
The month-3 churn spike is mostly the unflagged pricing change, concentrated in non-US cohorts where the new price hit without FX adjustment. Roll back FX-adjusted pricing for non-US, then re-measure cohort 12.
What changes for you
If two of three flag it, it is probably real. If they disagree, you see the disagreement, not a confident hallucination.
Charts, PDFs, screenshots, and CSVs all welcome. Each model picks up what the others miss.
The verdict pulls quotes back to the source so you can defend the read in the meeting.
Which mind to ask
Each model leans differently. Maya keeps the differences visible.
Best at qualitative read-outs, narrative analysis, and tying findings back to a thesis.
Strongest at numbers, tables, statistical reasoning, and code-driven analysis you can re-run.
Cuts through visual content, dense slide decks, and large image sets fast.
Where teams reach for it
Hear from three minds, get one verdict.
New accounts get $1 of trial credit. No card required to begin.