Rohit Patel

Rohit Patel headshot
Meta logo
Director, Generative AI
Meta

Presentation Title:

AI Evaluation 101:
Addressing Challenges to Real-World AI Applications

Presentation Summary:

This session looks at GenAI under the hood, breaking down how neural networks generate language. We discuss challenges to real-world AI deployment, and why evaluation is key to ensuring success. We’ll compare automated, judge-based and human-rated methods for assessing AI performance.

Attendees will learn how to build evaluations for their use-cases to assess AI vs. human performance on their tasks.

Picture of About | Rohit Patel

About | Rohit Patel

Rohit Patel is a Director on Meta’s Generative AI team, where he focuses on building the company’s next generation of AI models. Known for his ability to distill complex technical ideas into clear, accessible insights, Rohit is a sought-after voice in the fast-evolving world of artificial intelligence.

In addition to his leadership role at Meta, Rohit also founded and runs a research lab, QuickAI, which builds smaller, task-specific AI models for practical impact. He also advises Baypine, a digitally focused private equity firm, helping shape their AI strategy for acquisitions and across its portfolio companies.

Rohit’s unique blend of technical depth and business acumen is grounded in a career that spans enterprise consulting and product innovation. Before joining Meta, he advised CEOs of private equity–backed companies at PwC Strategy&, crafting analytics and data strategies to unlock enterprise value.

He holds a PhD in Game Theory from the Kellogg School of Management at Northwestern University, as well as Master’s and Bachelor’s degrees in Statistics from the Indian Statistical Institute. His article “Understanding LLMs from Scratch” has been widely shared by technologists and non-technical readers alike — an example of his gift for making AI concepts approachable and relevant.