Talp closes a pre-seed round at a $20 million valuation to simulate customer reactions with AI personas

Talp, a startup with Turkish roots, has closed a pre-seed round at a $20 million valuation, announced July 2, to build AI personas that simulate how customers will react to a product, ad or price before it launches. It was co-founded by CEO Baran Atas and Samet Alan, and has not disclosed how much it raised.

The round drew Formus Capital, Sunshine Lake Ventures, Aito Capital, the a16z Scout Fund and angel investors. The a16z Scout Fund lets people in Andreessen Horowitz’s network write small checks with the firm’s capital, a signal-driven bet rather than a conviction-sized one. Appetite has built around using AI to replace or supplement slow, expensive market research.

Talp’s pitch targets what it calls the “say-do gap,” the well-worn problem that people tell surveys one thing and then behave differently. By modeling synthetic customers, the company argues it can preview reactions faster than traditional panels and focus groups.

The premise is unproven. Simulated personas are only as good as the data and assumptions behind them, and whether AI-generated responses reliably predict real buying behavior is exactly the question the industry has not settled. Talp has not published accuracy results, and the pre-seed backs a bet, not a track record.

The startup enters a crowded and richly valued niche. Rival Aaru previously closed a Series A above $50 million at a reported $1 billion headline valuation on under $10 million in annual recurring revenue, a gap between price and sales that signals how speculative the synthetic-customer-research category still is. Whether Talp’s approach produces predictions clients trust enough to act on is what will separate it from the froth.