Profila
Sign up
The position
The consumer intent rail — where 6 billion people can say what they want, and brands pay to answer.
The pitch
The ask and the milestones behind it are in the deck. Download it, then talk to us.
// the_problem
Watch consumers, they perform. Stop watching, they tell the truth.
Every incumbent infers intent from behaviour recorded under observation. The AI underneath learns from the performance.
// the_solution
Give customers agency and privacy — they declare what they want, and from which brands.
Then they get exactly that. No surveillance. No guessing. The first place people say what they want, because they are finally in control.
// the_market
The money isn’t scarce. The signal is.
Inferred signal is watched, guessed, decaying. Declared signal is consented, fresh, appreciating.
Google $214B · Meta $196B — built on surveillance. We ask.
// the_flywheels
We land the brands. The brands bring customers. Customers talk — and it grows on its own.
B2B2C: brands invite customers → customers signal → brands serve → customers engage → brands convert.
Organic: customers vote for brands → search and discover → create new signals → attract new brands.
// the_platform
Three surfaces, one rail — live today.
Customers: vote and connect, signal, discover, engage.
Brands: listen, serve, scale — over OpenRTB 2.6 or MCP, never personal data.
Profila: growth, discovery, costs, and the data corpus underneath.
// the_business
Now the money. Customers get the content matches they asked for — paid placements that actually work, across every vertical.
Ad revenue today. And every served content match is scored — quietly building the data asset underneath.
// the_data
A dataset that captures what people chose, and has no public equivalent.
Verified records of what people chose matching their intent — not available elsewhere in consented, outcome-verified form.
A meaning model that adds value to any open LLM. The long game.
The proof
Every grant, paper and collaboration behind the eight years is published on Research Labs — including 42-True, the meaning-model paper.
Next step