Live from TWIMLcon! Operationalizing ML at Scale with Hussein Mehanna

800 800 The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)

The live interviews from TWIMLcon: AI Platforms continue with Hussein Mehanna, Head of Machine Learning and Artificial Intelligence at Cruise.

Starting at Facebook in 2012, with a machine learning department of just three people, to Google with an ever-expanding portfolio and now to Cruise, leading the trend of autonomous vehicles, Hussein has seen first hand what it takes to scale and sustain machine learning programs. At Facebook, a few early wins in the realm of infrastructure building set the stage for scaling via faster algorithms and trained models, and the ML team garnered much-deserved attention from the rest of the org. Soon, the entire Facebook organization could achieve a new level of ML scaling with all workflows shareable, reusable and discoverable through a search interface. This philosophy of sharing and openness has stayed with Hussein throughout his career. He was drawn to Cruise because of the interplay between applied research problems and the underlying tools and platforms, along with the immense capacity that the industry of autonomous vehicles has to push ML and AI to new limits of depth and scale. The complexity required to make decisions in such an unpredictable environment, combined with the critical need for safety and scale of data makes autonomous vehicles the sector of ML and AI to watch out for. Hussein breaks down the challenges (and joys) of working in the industry and his insight into analyzing scale when innovation is happening in parallel with the development. Listen to this episode to hear more of Hussein’s experiences at Facebook, Google, and Cruise, along with his thoughts on productivity being a “usability” versus “modeling” challenge and his prediction for the future of ML platforms!

About Hussein

From the Interview

Check it out

“More On That Later” by Lee Rosevere licensed under CC By 4.0

Leave a Reply

Your email address will not be published.