AI Evaluation Done Right: From Vibes to Trust
Building an AI prototype is easy. Proving it's ready for production is hard.
Most teams rely on "vibes", checking a few outputs manually and hoping for the best. But when you need to deploy to thousands of users, you need a real system to measure quality, reliability, and risk.
AI Forward, the Bereisheet Tech AI community, is hosting a practical session on how to solve this. We are bringing together 3 perspectives: broad consultancy, frontier research, and industry.
No hype, just the tools and frameworks you need to move from "it works on my laptop" to "it works in production."
🗓️ Agenda
18:00 | Reception & Networking
18:30 | Opening Remarks
18:40 | Part 1: The Foundation – The Eval Tech Stack & Terms Speaker: Lior Kanfi, CEO & Founder of Tikal
Before we can fix the problem, we need to speak the same language. Lior sets the baseline for the evening by defining some essential terms and mapping out the "Eval Tech Stack," the specific tools and workflows required to turn evaluation into a reliable process.
19:00 | Part 2: The Science – How to Measure Quality
Speaker: Ofir Press, Postdoctoral Researcher at Princeton University and the creator of SWE-bench (used by OpenAI & Google).
Ofir will explain how to build a valid benchmark for your specific project and how to use it to choose the right model.
19:30 | Part 3: The Real World – From High Stakes Evaluation to Production Code
Speaker: Matan Barak, Head of Legal Product at Verbit With live demonstration by Shay Yahal, AI Eval Builder
A deep dive into evaluation in the legal industry, where small errors can have big consequences. This session combines real world challenges with a hands on demonstration of how to turn theory into working code.
20:00 | Meet the AI Community Good food, drinks, and networking with fellow builders.
🗣️ Language: Hebrew
The difference between a cool demo and a scalable business is evaluation. Join us to find out if your AI is ready for the real world.
🙏 Thank You to Our Host Pearl Cohen, an international law firm with offices in New York, Boston, San Francisco, Tel Aviv, London, and Munich. Representing innovation driven companies and their investors, they serve clients worldwide, from Fortune 500 companies to leading academic institutions and promising startups. With professionals coming from a wide variety of academic and professional backgrounds, their team is your partner in the business of innovation.
All event payments are considered donations to Bereisheet and are non-refundable.