Sign In

Ronny Kohavi

Vice President and Technical Fellow - Data Science, Engineering - AI, Machine Learning, Controlled Experiments - Ex-Airbnb, Ex-Microsoft, Ex-Amazon

Ron Kohavi is an accomplished executive leader and technological innovator with expertise in controlled experiments, machine learning, personalization, and AI.

He has served in leadership roles at top tech companies like Airbnb and Microsoft, where he accelerated software innovation through trustworthy analysis and experimentation, leading teams to discover new ideas and drive incremental revenue.

With a background in Computer Science, including a Ph.D. from Stanford University, Ron Kohavi is a recognized thought leader in AI and Machine Learning with numerous citations and impactful published work.

He is a Quora top-viewed writer in A/B testing, a prolific speaker at conferences, and holds 15 patents in areas like Machine Learning, Visualization, and Experimentation.

Ron Kohavi's contributions to the industry have revolutionized accuracy estimation, feature selection, and controlled experiments, shaping best practices implemented across organizations.

Educationally, he excelled in his studies at the Technion in Israel, graduating Summa Cum Laude with a BA in Computer Science and passing his Ph.D. qualifying exam with distinction at Stanford University.

His career spans key roles at companies like Amazon.com, Blue Martini Software, and Silicon Graphics, Inc (SGI), showcasing his diverse experience in tech leadership and innovation.

Highlights

Aug 14 · linkedin

...maybe the biggest educational contribution that could be made to the creativity of people is to persuade them that they shouldn't worry about making mistakes. This will be inevitable.  -- William Shockley, nobel prize winner and inventor of transistor

Great line from the history of Silicon Valley In the Beginning, narrated by Leonard Nimoy: https://lnkd.in/gUF28haU

In my world of A/B testing, failures rates of 85% are common. Fun discussion on whether the 92% failure rate at Airbnb search implies that one should ask why "you keep making consistently poor guesses": https://lnkd.in/gegXfZeD

Here is the full section of the interview:

It is my own experience that to do creative work, one must overextend oneself, one must count on falling on his face, on getting into difficulties. One must learn from these failures and not be stopped by them. If one is taught that everything is neat and orderly, and one never gets into a mess when trying to do anything new, then you will be so conservative that I don't think  you'll break new ground. I think the big contribution that can be made, maybe the biggest educational contribution that could be made to the creativity of people is to persuade them that they shouldn't worry about making mistakes. This will be inevitable.

Jun 2 · linkedin

The paper, False Positives in A/B Tests, was accepted to KDD 2024:        https://lnkd.in/gVjj4Wpv

We are publishing our nearly final draft for feedback for two reasons:

1. The reviewers wrote that some sections are “challenging for readers without a strong statistics background.” We have incorporated more explanations, two additional figures, and we would love to get more feedback before June 12th about things we could clarify, and improvements we can make, or anything we can fix.

2. We think this is an important and under-appreciated topic. The meta-reviewer wrote: "The paper has the potential for broad impact on experimentation practices across many domains…The novel sequential testing procedure is both theoretically rigorous and practically relevant." One reviewer wrote: "The great details on calculations, rigorous analysis, practical recommendations and real-world examples make it a must read for researchers and practitioners alike."

Nanyu Chen #abtest, #experimentguide

Ronny Kohavi
Add to my network

Location

Los Altos, California