AI Engineer

About the Team

Motives AI team builds the models and systems that let us see and understand whats happening on the road and in the field, across cameras, sensors, and millions of hours of fleet data. Our work spans model development, on-device deployment, and the cloud infrastructure that turns raw signals into products drivers and safety teams rely on every day.

About the Role

Were looking for an AI Engineer to work across the full stack: training and evaluating models, then getting them running in the real world and building the systems around them. Youll partner closely with platform, product, and firmware teams to ship your work.

Youll own the work end to end. You might spend one day improving a model, another building the pipeline that feeds it, and another in production tracking down why somethings off.

In this role, you will:

• Research and develop machine learning models for perception and understanding problems across visual, audio, and other real-world signals.
• Explore how specialized models and larger general-purpose models can work together in production systems.
• Design data, training, and evaluation approaches that hold up under real-world conditions, not just in a notebook.
• Study model behavior, robustness, and failure modes across data, deployment, and operating conditions.
• Integrate and validate new capabilities in real-time or resource-constrained systems.
• Work with firmware, platform, and product teams to turn research into working systems.

You might thrive in this role if you:

• Understand deep-learning fundamentals: architectures, training dynamics, evaluation design, and data quality and annotation.
• Think in systems: latency and memory budgets, failure modes, distributed pipelines, and on-device constraints.
• Take end-to-end ownership across the model, edge, and cloud boundaries, and stay comfortable with ambiguity along the way.
• Thrive in fast-paced environments and can rapidly iterate from experimentation to production.
• Have hands-on experience using agentic development tools and AI-assisted coding as a core part of how you build and ship, not as a side experiment.
• Are proficient in Python and PyTorch, with working ability in at least one systems language (e.g., C++ or Go).

Strong candidates may also have:

• Experience building LLM or agent-based applications including tool use, retrieval over domain data, and evaluating non-deterministic systems.
• Experience with on-device or embedded ML (SNPE, TensorRT, TFLite, ONNX Runtime, or similar).
• Experience with video understanding, VLM/embedding models, or large-scale retrieval.
• Experience working with large-scale video or telemetry data, including annotation pipelines and observability tooling (e.g., Snowflake, Redash).
• Experience with distributed training or experimentation frameworks (e.g., Ray/Anyscale) and building rigorous evaluation harnesses.