Building ML infrastructure and teams
Currently building ML Feature Engineering Experience at Wolt & DoorDash.
Hiring Staff ML Platform Engineers!Environmental Science → Data Science → ML Engineering → Engineering Leadership
My journey into data began with remote sensing and land cover detection using segmentation CNNs in 2015. I became fascinated with the idea of "planetary-scale computing" for stewarding the earth-system but frustrated with academia. Instead I chose to deepen my skills first as a GIS developer and then later Geospatial Data Scientist.
Early in my career I worked a lot with satellite and drone imagery as well as different kinds of 3D scans. Working for a consultancy, I helped clients with use cases from bark-beetle infestation detection to construction site monitoring and 3D object detection. I went from building custom PyTorch implementations of novel 3D architectures from research papers to leading the launch of our 3D classification platform.
Working with large unstructured datasets came with unique challenges for data processing and I gravitated towards building infrastructure so that my team could ship actually deployable solutions. I expanded my skillset to become effective with Docker, Kubernetes, and different cloud providers and enjoyed that a lot.
In 2021 I joined Wolt's data team and helped build the company's in-house ML platform and team from the ground. I saw the company grow from ~200 to more than 1000 engineers and our ML user base from less than 10 to more than 50 people, enabling them to train and serve 100s of models accross multiple clusters.
In 2025 my team joined DoorDash's ML Platform to cover Feature Engineering in a global team that supports both companies. The scale is extremely ambitious and requires rethinking how we do feature engineering to make it a truly seamless part of model iteration.
Coming soon...