Office Hours

I forgot how fun small group networking dinners are!! Thanks Office Hours, Avenga & CXOsync for inviting me to these very special dinners. It was incredible to meet leaders in Healthtech (👋🏽 Abdel Mahmoud, MD Rob Brisk Ernest Lim Xiao Liu ) and other industries (👋🏽 Christina Habib Anthony Fernandes Flavio Vincenti )! 💡Based on what I heard from other attendees, my takeaways: ➡️ We should be proud at #UHL & #UHN for delivering value from use of #AI & #AgenticAutomation at a scale no one else has figured out how to do. Not in the NHS, neither in the private sector in large banks, tech scales ups and retail. ➡️ It’s truly so refreshing to connect with others outside of healthcare (and the NHS) to share learnings. There are loads we can learn from each other! ➡️ It’s not so easy to set up #AILabs and #EmergingTechCapabilities in large organizations. Other large organizations (private and public) are struggling to figure out how to set up for scale & long term sustainability. Having personally built one from scratch at UHL/UHN, servicing a 30,000 people organization, I can say it was probably one of the most challenging work in my career to date, but I’m proud to say we are “almost” becoming a well-oiled machine. Looking forward to making time to attend more of these this year!

Been a bit inactive here, but wanted to share that I joined Office Hours as part of the Client Solutions team a while ago. It’s been really good so far, awesome people, a crazy fast-moving environment and a lot of learning along the way. Super happy to be working with such a great team. Excited for what’s ahead 🚀 --- Pratheek Velamuri Aditya Kalwani Subhash Movva Natalie Grover Naina Bindra Archita Agarwal Devmini W. Ashton Guss Noah Katz Ansh Tanna

This week Massive and AI Circle brought together leaders from the most cutting edge labs in the world! It was a great opportunity to learn with engineers and researchers from companies whose products and data we use everyday: Instagram, Google DeepMind, Microsoft, Amazon, Anthropic, Braintrust, Appen, Office Hours, Cohere, Massive, and more. Grateful for Albert Chun and Suzanne Rabicoff from AI Circle in helping bring this to fruition. Thanks to Adam Sah (Zero Capital) and Leesa Lee for hosting us in your beautiful home! Also, thanks to Adam for remembering to take photos. I snuck into this one that was supposed to be candid 😎

Last night, it struck me: our members in the room have a great deal of influence over the future advancements of our industry. We had members working in data collection, data annotation, pre-training, post-training, evals, inference, applied AI, trust and safety, deployment at the most cutting edge labs and largest enterprises in the world like Instagram, Google DeepMind, Microsoft, Anthropic, Braintrust, Appen, Office Hours, Massive, and more. I learned so much last night: - agent harnesses for long-horizon tasks across multiple context windows that spans across hours and days have replaced agentic workflows in many instances. - trust and safety framework for model need to be broad enough to address geopolitics, global threats, and cyber security and it gets as practical and on the ground as addressing data leakages, catastrophic errors, and military defense. - what high quality tokens look like from a pre-training perspective and how often startups submit terrible samples with incoherent formatting and inconsistent directories. - what AI deployment looks like for the most highly regulated industries like financial services, especially in the financial capital of the world New York City Major thanks to our member Jason Grad and the team at Massive for co-hosting this event with us. I'm especially grateful for Adam Sah, founder of Zero Capital, and Leesa Lee for opening up their beautiful home.