Senior / Staff Software Engineer, ML-based Controls

Waabi

About The Position

Design and develop data-driven and machine-learned approaches to vehicle control problems, bringing modern ML to a domain traditionally solved with classical methods. Develop learned models of vehicle behavior and dynamics, and integrate them into the closed-loop simulation. Apply machine learning to improve how the controller adapts across vehicles and operating conditions. Be part of a team of multidisciplinary Engineers and Research Scientists using an AI-first approach to enable safe self-driving at scale. Own problems end to end, from conceptualization and offline experimentation through simulation and on-vehicle validation. Build the data pipelines, evaluation metrics, and tooling needed to measure whether a learned approach outperforms the classical baseline. Participate and share ideas in technical and architecture discussions, helping define how learning and classical control coexist in a safety-critical stack.

Requirements

MS/PhD or Bachelors degree with a minimum of 4 years of industry experience in Robotics, Controls, Mechanical/Electrical Engineering, Computer Science and/or similar technical field(s) of study.
Demonstrated depth in control theory and dynamic systems (e.g., MPC, optimal control, state estimation, system identification, kinematic and dynamic vehicle modeling).
Hands-on experience applying machine learning to a physical system, with real hardware in the loop rather than simulation alone.
Production-quality coding skill in Python and C++.
Experience with deep learning frameworks such as PyTorch.
Solid problem solving skills using linear algebra, optimization, statistics & probability.
Ability to rapidly prototype and test new algorithms.
Ability to design experiments that prove whether algorithms work.
Open-minded and collaborative team player with the willingness to help others.
Passionate about self-driving technologies, solving hard problems, and creating innovative solutions.

Responsibilities

Design and develop data-driven and machine-learned approaches to vehicle control problems.
Develop learned models of vehicle behavior and dynamics, and integrate them into the closed-loop simulation.
Apply machine learning to improve controller adaptation across vehicles and operating conditions.
Own problems end to end, from conceptualization and offline experimentation through simulation and on-vehicle validation.
Build data pipelines, evaluation metrics, and tooling to measure performance against classical baselines.
Participate in technical and architecture discussions to define coexistence of learning and classical control.