Suggestions
David Purdy
Experienced Data Science and Technology Leader
There appear to be two different people named David Purdy:
David Purdy, CEO of Palladio AI
This David Purdy is the CEO and co-founder of Palladio AI, a startup in the San Francisco Bay Area.1 He previously held leadership roles in data science and machine learning at companies like Apple, Uber, and Theranos.1
David Purdy, CEO of Gastrograph AI
The other David Purdy is the CEO of Gastrograph AI, a company based in New York City.23 He has over 15 years of experience in technology leadership roles at companies like Cantina and SwiftIQ.2 He is also the founder and executive director emeritus of the Tea Institute at Penn State University.4
The LinkedIn profile at https://www.linkedin.com/in/davidpurdy appears to belong to the David Purdy who is the CEO of Palladio AI1, not the David Purdy who is the CEO of Gastrograph AI.2
Highlights
I am delighted to share that I've founded Palladio AI. More to come!
It was great to be back at Uber last week to talk with the Machine Learning Platform team about the founding of the team and its platform, Michelangelo. I was particularly grateful to be hosted by Eric Chen (pictured), one of the first two engineers who joined me on the project in early 2015.
Michelangelo’s impact and adoption are well documented, with machine learning models involved in many applications, including ETA estimation, fraud prevention, safety, and much more. Dara Khosrowshahi, Uber’s CEO, has noted that machine learning is at the heart of many of Uber’s services. Throughout every trip, every session, and every day, this platform is powering models to make meaningful decisions across hundreds of applications. Without the team and this platform, Uber would be far less agile, robust, and efficient. In fact, Uber’s investments in machine learning were noted in its S-1 paperwork when it filed for an IPO.
Michelangelo was one of the big, bold bets Uber made in technology. When I proposed an enterprise platform, Uber had just 1 machine learning model in production. However, the future was clear: machine learning would become pervasive. Leading ML adoption is why I joined Uber. Before Uber, I’d built and deployed hundreds of models, yet nothing compared to Uber for the potential real-world impact of ML.
In the talk, I covered many of the early decisions, including the origins of the name, the buy/build decision, the lessons I brought to Michelangelo, and those lessons I took from the experience. For me, Michelangelo was a terrific opportunity for professional growth, leading it from vision and mission to architecture, roadmap, and user advocate. On a personal note, launching the Michelangelo team clarified the difference in mindsets of an employee holding shares versus a shareholder building a company.
I also learned something from giving the talk. I’ve long focused on solving problems and building teams that solve problems. This experience reminded me of the importance of sharing lessons learned so others can develop their toolkits for solving problems. I look forward to sharing more lessons learned throughout my career.
I am super proud of our investment in Michelangelo and am exceptionally impressed by the team’s achievements. The team has evolved the objectives and systems far beyond the early goals. Their craftsmanship and innovation have certainly earned the name. Thanks to everyone on the team. A special thanks to Eric Chen, Kai Wang, Min Cai, and Joseph Wang, with whom I’ve had so many great discussions and joint efforts over the years, as they've led Michelangelo to ever greater impact.
If you’d like to see the latest on Michelangelo, I encourage you to check out Min Cai’s talk at Systems@Scale a few weeks ago: https://lnkd.in/gKkjcYTg
#michelangelo #uber #mlops #mlplatform #uberMichelangelo #machinelearning #platforms
