• Tech Lead for Gemini spatial understanding

• Tech Lead for vision and language (VLM) object detection models across Google & Alphabet

• Gemini core: vision, large scale pre-training (LSP), post-training

• Image Understanding

• Mobile Vision

• C++ mentor

• Python mentor

• Tech Lead for production computer vision search engine based on embedding retrieval (Scala, Python, C++) including GPU-enabled microservices on Kubernetes.

• Research into image embeddings for E-Commerce applications (TensorFlow, Caffe)

• Tech Lead for automated training pipeline to increase team velocity.

• Object detection deep learning (Theano, TensorFlow, Caffe)

• Automated training of models using AWS EC2 spot instances, S3, and SQS.

• Deployed models on an NVIDIA Jetson TX1 to support object detection predictions at the edge.

• Research into predicting ancestry from genetic data. Over 10,000 users in the first three months.

• Also owned the database, server, and backend data ingestion pipeline.

• Statistical modeling of chemical environmental, demographic, and health data for expert witnesses and trial exhibits.

• Owner of R and C++ packages for chemical mass transfer.

• Managed teams collecting and analyzing data related to emissions reconstruction and evaluation of cleanup remedies.