Technical Lead Manager, Auto Labeling

The Auto Labeling team develops trained algorithms, tools and infrastructure to provide labels that can be used by Perception, Motion Planning, and other ML teams for training and evaluation. The team works closely with manual labeling efforts and infrastructure to create a data centric ecosystem needed to develop real time, safety critical ML models for autonomous driving.

We are looking for a technical lead manager for the Auto Labeling team. Candidates should have a strong background in architecting, training, and deploying Deep Learning Models for Perception or Prediction, and experience in data centric model development. They should have a mission-driven mindset to deliver a product, which entails significant cross-functional interaction, and customer-centric obsession to deliver what ML teams at Stack AV need.

What We're Looking for:

• Build and lead a team of engineers to develop state of the auto labeling tools for self driving vehicle systems.
• Set the technical direction for the team, and work cross-functionally to develop safe systems.
• Define the strategy and roadmap for automatic data labeling of autonomous driving datasets.
• Drive execution of the team’s roadmap, review their code and SW architecture, and deliver features in lockstep with key partners.
• Define labeling and dataset requirements, in close collaboration with manual labeling teams and ML modeling teams across Perception, Prediction, Planning, and other autonomy teams.

What Success Looks Like:

• BS/MS/PhD in Computer Science or related field, or equivalent industry experience (5+ years of experience).
• Experience with strong, hands-on leadership of a team of engineers.
• Understands state of the art in pseudo labels, self supervised learning and unsupervised learning.
• Familiar with techniques such as vision transformers, large foundation models, models such as Segment Anything Model.
• Experience with architecting, training, and deploying Deep Learning Models.
• Familiarity with offline perception models and acausal information for auto labeling.
• Fluent in python, C++ fluency is a plus.