Early access · intent data from AI-powered apps
People tell AI apps what they’re planning long before they search for it. Velmax is building the consent-first network that turns that activity into audience and intent data your marketing team can put to work.
01Data products
Four products, one consent-first signal network. Each is designed to ship with a plain-English definition, a refresh cadence and the consent basis behind it.
AAudience segments
Audiences defined by real planning moments — from “moving in the spring” to “shopping for a first EV” — built to activate in your DSP, CRM or clean room.
BIntent signals
Topic-level interest aggregated by region and hour, designed to show demand building before it reaches search volume.
CTrend panels
Indexes of rising and falling topics across the network, with category and regional cuts your analysts can model against.
DCustom cohorts
Write the audience you want in a sentence. We’ll model a cohort against the network and come back with a sized, documented definition.
02Catalog
The first segments on our roadmap. Each will ship with its definition, sources, refresh cadence and consent basis. Drag the ring, then open any card for its draft spec.
TravelSEG-0142
AutoSEG-0317
HomeSEG-0409
FitnessSEG-0521
FinanceSEG-0633
FoodSEG-0718
EducationSEG-0806
PetsSEG-0912
RetailSEG-1024
GamingSEG-1107
Small businessSEG-1233
BeautySEG-1350
Planned segments · draft specs, subject to change
03Method
Partner apps send event-level signals through a lightweight integration — only for users who opted in under the app’s own privacy notice.
Direct identifiers are dropped or replaced, and free text is classified on the device. The topic leaves the app; the words never do.
Signals are mapped to a topic taxonomy, scored for intent strength and recency, and only aggregated into cohorts above a minimum size.
Segments, feeds and panels are built to land where your team already works — S3, Snowflake, BigQuery or an API — on the cadence you choose.
03Method
A four-stage pipeline designed so that the most sensitive part of every interaction — what someone actually typed — never reaches us.
Partner apps send event-level signals through a lightweight integration — only for users who opted in under the app’s own privacy notice.
Direct identifiers are dropped or replaced, and free text is classified on the device. The topic leaves the app; the words never do.
Signals are mapped to a topic taxonomy, scored for intent strength and recency, and only aggregated into cohorts above a minimum size.
Segments, feeds and panels are built to land where your team already works — S3, Snowflake, BigQuery or an API — on the cadence you choose.
04Trust
Marketing data is only worth buying if you can defend where it came from. These are the rules every Velmax dataset is built under, starting with the first.
"app""trip-planner"
"consent""opt-in · v3 · 2026-09-14"
"user""maya.r@example.com"
"device_id""A1F3-77C2-90DE"
"prompt""flights to Lisbon in May under $700?"
"location""39.9526, -75.1652"
"topic"null
"intent"null
}
Every signal comes from an app user who opted in. No scraping, no purchased lists, no inferred consent.
We never receive message text, names, emails or phone numbers — so we can never sell them.
Segments below our aggregation floor are never built, sized or released.
We don’t model health conditions, sexuality, religion, politics, children or precise location.
Consumer opt-out requests are honored in every future build and delivery.
Every dataset will ship with its definition, sources and consent basis, ready for your legal and privacy review.
05Founding partners
Velmax is choosing its founding app partners: teams who want to turn aggregate, consented intent into revenue — with consent language users can understand, on-device classification, and a clear share of every dataset their app contributes to.
06Contact
Share the audience, the channel and the timeline. Early-access requests shape which segments we build first — and you’ll hear the moment they’re ready.