One Year Beyond Amazon

I left Amazon a year ago.? My departure ended up being more public than I expected, perhaps in part because I had staunchly defended the company’s response to COVID-19? just a few months earlier.??

I had been a deeply loyal, missionary Amazonian. I spent a huge amount of effort helping transform Amazon’s home-grown core eCommerce engine into a system capable of supporting Amazon’s ever-expanding vision of being an “everything store”.? I oversaw the technical readiness for the first Prime Day shopping event and stayed up that night and the next to keep the systems running under a shopping surge well beyond anything we had anticipated.? I led Amazon’s Project Rolling Stone to migrate the company off Oracle databases.? I helped bootstrap the early Prime Air program by hiring 19 interns the first summer to explore all avenues of air-based delivery.? Then I took on the immensely challenging job of speeding up Amazon’s deployment of new robotic systems to improve safety and help increase the capacity of the network both in storage and throughput to enable one-day shipping.? When COVID-19 hit, I pulled together diverse teams across operations and redeployed many in my own Robotics organization to explore and invent ways to ensure Amazon was able to apply its speed and technology to keep associates safe.??

I am proud of what I did at Amazon.? But companies change.? I could tell Amazon was changing.? I didn’t know Jeff Wilke would retire.? I didn’t know Jeff Bezos would step out of the CEO role.? I didn’t know Charlie Bell or Wei Gao or dozens of my peer VPs would leave too.? But I knew something was different.? I longed for the earlier, more egalitarian days.? I yearned for the opportunity to help build from scratch.????

Prior to Amazon, I had spent 10 years in hyper-growth startups at Netscape and Tellme Networks.? I knew I needed to get back to that energy, that pace, that sense of togetherness.

As soon as I reconnected with Alex Wang at Scale AI and he shared his vision to become the best AI company in the world, not through shortcuts, but by really grinding to help solve the biggest bottlenecks in getting AI to product… namely data quality, experimental velocity, and world-class models… I was hooked.??

In the past year, we’ve doubled the engineering and product organization.? We’ve launched our Nucleus product for dataset management and debugging, our Document AI product for template-less data extraction, our Scale Rapid experimentation platform, our Catalog quality offerings, our Mapping product, our new Frameless Lidar labeling product, and built our capability to help the government bring AI to life in service of the nation.? We’ve invested substantially in our already industry-leading ability to deliver high-quality labels.? We’ve invested in operational efficiencies to deliver that quality more reliability, predictability and at lower cost.? We’ve invested in our platforms and systems.?

Entering year two, we’re expanding our offerings further to help customers combine their product data with Synthetic Data to drive ML performance faster.? We’re expanding our operational capability and technical toolset to accelerate Data Collection, particularly for AR/VR use-cases but well beyond that as well.? We’re building out our Content Understanding and Model CI capabilities.??

And we’re building the mechanisms to ensure a culture that is fair and loyal to our amazing and deeply missionary employees.? The last is what makes me most proud.?