Laszlo Korsos

I work at the intersection of AI, marketplace, and applied science. Today I lead AI/ML, Data Science, Research Science, and ML Platform at Attentive — a 25-person organization owning end-to-end modeling and inference for 8,000+ enterprise brands. I'm also co-founder of Multiply Talent, the largest boutique executive search firm serving private equity for AI, ML, and data leadership.

I've spent 15+ years building the algorithms, the teams, and the businesses behind systems that touch hundreds of millions of people daily — from the dynamic pricing code that priced every Uber trip worldwide and generated $300M+ in incremental annual revenue, to the SMS frequency model that unlocked a ~$6B annual revenue opportunity at Attentive. My academic background is in probabilistic machine learning and Bayesian inference. I finished my PhD at the University of Chicago in a record 2 years and 9 months, built one of the world's Top 500 supercomputers as an intern at a global investment fund, and used the proceeds to finance expeditions to summit some of the world's highest mountains.

Outside of work, I'm a mountaineer, endurance athlete (50+ marathons/half marathons, an ultra, an Ironman 70.3), and an enthusiast of Napa Valley viticulture. I've also sung backup for Kenny Chesney, Ashanti, Renée Fleming, Darius Rucker, and Joss Stone.

Experience LinkedIn →

2024–Present

Attentive

Head of AI/ML, DS, Research Science & ML Platform · First Data/Applied Science IC · 2024–Present

Promoted three times in 18 months — joined as the first applied science IC; now lead a 25-person org spanning AI/ML, Data Science, Research Science, and ML Platform, owning end-to-end modeling and high-volume inference for 8,000+ enterprise clients including Target, H&M, Crate & Barrel, Wayfair, Skims, Alo Yoga, and Crocs. Built the SMS industry's first frequency optimization model end-to-end, unlocking the ~$6B annual revenue opportunity that now underpins most of the company's product roadmap.

2019–Present

Multiply Talent · f.k.a. HeadRace

Co-Founder · 2019–Present

Founded a Data/AI-focused recruiting business; bootstrapped to $1M+ in revenue as a one-person side-hustle, then raised $6M Seed led by Greylock with Susa Ventures and Breyer Capital. Scaled into a venture-backed marketplace; won Slack's "Visionary of the Year" alongside OpenAI, Rivian, Shipt, and Spotify. Brought in a dedicated CEO with leadership experience at Korn Ferry and Heidrick & Struggles to run the day-to-day, and relaunched as Multiply Talent — now the largest boutique AI, ML, and Data executive search firm serving private equity.

2020–21

Realtor.com

SVP of Data & Analytics · 2020–21

Led data and analytics across the entire business — Product, Engineering, Sales, Performance Marketing, and Brand Marketing. Inherited a team of 35 and grew it to 60+ in under a year. Drove the shift from centralized analytics to embedded analysts partnering directly with product and business teams.

2019

Flexport

Senior Director, Head of Marketplace & Pricing DS/ML · 2019

Led the team that created Flexport's industry-leading pricing and optimization technology across procurement, margin, routing, and fulfillment.

2014–17

Uber

Co-Founder of Marketplace Org · Founder of Marketplace Data Science · 2014–17

Joined as one of Uber’s first 100 technical hires. Promoted within 2 months to found and lead the teams behind Uber’s core marketplace algorithms — Surge Pricing, UberPool matching and routing, dispatch optimization, and anomaly detection — pricing every Uber trip worldwide. Led development of H3 hexagon-grid dynamic pricing, generating $300M+ in incremental annual revenue. Co-founded the broader Marketplace Team (now 1,000+ people). Co-inventor on 3 patents and 7 patent applications. Founded what Travis Kalanick called Uber’s “math department.” Cold-recruited 5 tenured professors from Columbia, Yale, Cornell, and UCLA.

2013–14

Goldman Sachs

Associate, Interest Rate Product Strats · 2013–14

Structured interest rate derivative deals. Implemented the firm's first Bayesian statistical learning methods for asset pricing, hedging, and forecasting.

2009–10