▶ Watch ↗15:51
Conference affiliation: Head of Data, Digital, and AI · Apollo Global Management · 2026
Adam Nahari is a Boston-based Data, Digital & AI Operating Partner in Apollo’s Portfolio Performance Solutions group. He oversees two teams. The data science and research group helps investment teams find opportunities, develop investment theses, and test their assumptions during diligence. The portfolio value creation group works with companies on operating decisions that affect their financial performance. Before applying data to investment and operating decisions, Adam studied what routine hospital records could reveal about emerging disease. Apollo
Adam earned a B.A. from Harvard University and worked as a research scientist in Harvard Medical School’s Predictive Medicine Group at Boston Children’s Hospital. In a study published in 2022, he and Ben Y. Reis analyzed five years of pediatric lab records to see what routine tests could reveal when disease-specific testing was scarce. They looked for shifts in abnormal results across the hospital and compared patterns in individual patients with known diagnoses. Some combinations of routine tests were associated with both COVID-19 and multisystem inflammatory syndrome in children, with stronger associations for the inflammatory syndrome. They proposed using those patterns to help prioritize patients for further testing. Pediatric lab study
In a separate study with John S. Brownstein and Reis, Adam helped design the study and conduct the statistical analysis. The researchers compared aggregate firearm-related internet searches with background checks across all 50 states. The two measures moved together across states and over time. Search activity could provide a timely supplement to background checks, which are themselves an indirect measure of firearm sales. Firearm search study
Adam joined Berkshire Partners in 2020 as vice president of data and analytics. He built data science and AI capabilities and worked with company managers to improve commercial performance and digitize core processes. He has described the variety as a reason he chose private equity. Moving among companies and industries changed both the questions he could ask and the data available to answer them. Apollo · Interview on the move into private equity
As an MIT System Design and Management Fellow, Adam earned an M.S. and used his 2023 thesis to test whether changes in companies’ Google review ratings could help predict subsequent stock returns. He simulated buying or shorting shares as ratings moved, varying both the period over which he measured the change and how long he held each position. Results varied by industry and time horizon. Bets on rising ratings fared better than bets against falling ratings, although Adam cautioned that the market’s substantial rise during the study period complicated the comparison. MIT thesis
As chief data and AI officer at Pinegrove Venture Partners, Adam led analytics for finding investment opportunities, assessing deals during underwriting, and managing investments after closing. In 2024, he and his MIT thesis adviser, Dimitris Bertsimas, argued that AI made data from outside a company more useful at each of those stages. They treated the choice of which external data to collect and use as an investment decision in its own right. Apollo · Harvard Business Review
Adam argues that general-purpose AI can save time across an industry. An individual company has a better chance of keeping the gains when it redesigns a recurring process around its own data and the people who make the decisions. In his cargo-airline example, the team targeted crew costs by giving schedulers a tool to compare feasible assignments by cost before the scheduling deadline. AI Engineer recording · Talk companion
For Adam, the funding decision starts with a line on the income statement, an operating owner who can change the process, and a baseline finance can use to measure the result. He wants finance and operations to agree before the pilot on how much improvement and adoption would justify continuing and when to stop. The decision to expand also has to account for the full cost of running the system at scale. Talk companion
▶ Watch ↗15:51
Loading more talks and highlights…