About Probat

We talked to hundreds of marketing leaders before building Probat. From brands to teams at eBay and Meta, one answer kept coming back: a page made for the person who clicked beats any A/B test.

Probat AI (probat.ai) makes that page for every ad, audience and account, and proves what it earns.

Founded 2025 · San Francisco · Probat AI Inc., a Delaware company

The team

  • Portrait of Sari Abu-Hamad in a navy blazer, white shirt and glasses

    Co-founder

    Sari Abu-Hamad

    CEO & Co-founder

    Sari studied computer science at UCLA, where he led projects for HP Tech Ventures and Visa in Bruin Ventures, the university’s venture and consulting club. Before Probat, he grew online communities in the crypto space to more than 50,000 members and spent years optimizing conversion for his own six-figure trading card business. At IBM he sold data and AI software: in his first quarter he hit 200% of quota and was named MVP across the Americas sales organization, then went on to manage 68 large enterprise accounts. There he watched a single page change cut inbound leads by 30%, which is why Probat tests every page it builds. At Probat he writes code across the product and works with every customer directly.

    Sari on LinkedIn
  • Portrait of Eric Choi in a dark suit and tie under brick arches at UCLA

    Co-founder

    Eric Choi

    CTO & Co-founder

    Eric studied computer science at UCLA and left a master’s at Columbia, where he researched how to protect AI agents from prompt injection, to build Probat full time. Over four summers at Google he worked on Spanner and BigQuery ML and built the AI agent that became Gemini’s Google Contacts integration. He also interned at Rockset through its acquisition by OpenAI. At Probat he builds the engine that writes and tests each page, the editor, and the system that learns from every test.

    Eric on LinkedIn
  • Portrait of George Sadler in a dark blazer and striped shirt

    Advisor

    George Sadler

    Advisor

    George led global marketing sciences at eBay, where his team ran thousands of experiments a year across email, search, social and loyalty. Those tests showed him how much more a personalized experience converts than a generic one. He then spent six years at Adobe running B2B revenue operations and account-based personalization, and led revenue operations at ID.me and Simpro. He has guest lectured on experimentation at scale in UCLA Anderson’s business analytics program.

    George on LinkedIn

See what we’ve built for our customers, read reviews of Probat, or read the blog.

See it on your own ads.

Book a demo