Product
AggregateEvery scattered source into one canonical timelineDetectReach decay, VOC spikes, competitor moves, stockout riskEvidenceThree-way attribution, hypothesis testing, hallucinations strippedActCreator tier focus, budget reallocation, new creative angleOptimizeContract acceptance, outcome ledger, control-group guardsCase Study
16 weeks after onboarding: measured vs control baseline
Case Study
Where it startedHow the engine ran itAsk it anythingThe 16-week resultBaseline & methodAll case studiesCase Study
16 weeks after onboarding: measured vs control baseline
No seats, no tiers, no hidden line items. Your bill is what the engine actually consumed in tokens, plus a transparent 20% service fee.
Onboarding & setup
Billed on real consumption
How billing worksWe pass model costs through at par; the 20% covers operations, support and continuous tuning. You can always account for every dollar — the same principle as every number in the product being traceable.
Yes. Onboarding, history backfill and the Day-0 audit are free — they consume almost no tokens. You see what the engine can find on day one, then decide whether to continue.
No minimum. Monthly settlement, stop any time; billing stops when you stop, and your data is returned or destroyed per agreement.
The bill itemizes every investigation's token consumption, matched to the evidence IDs in your reports — you can drill into exactly which sweep or evidence chain spent what.
The agent watches alerts and follow-ups around the clock; human strategists respond in depth during working hours, with an SLA for urgent items. Direct Slack / Feishu channel, no ticket system.