Principal Machine Learning Engineer, Data Mining

Mission Summary:

At Motional, we're transforming how autonomous vehicles discover critical intelligence hidden within petabytes of multimodal sensor data. Our next-generation autonomous driving stack depends on finding the rare edge cases, long-tail scenarios, and model errors that matter most. Omnitag, our ML-powered multimodal data mining framework, is the engine that powers this discovery.

As a Principal Machine Learning Engineer, you will serve as a foundational technical leader shaping the long-term, multi-year vision for our foundation model post-training ecosystem. While training foundation models from scratch is on our horizon, your immediate mandate is to architect the systems that adapt, align, and optimize massive multimodal models for our autonomy stack.

Armed with a strong sense of product, you will ensure our technical investments directly translate into downstream value, accelerating edge-case discovery, reducing engineering costs, and improving the final autonomous driving product. You will tackle highly complex, ambiguous problems, creating clarity and defining objectives that pay future dividends for the entire company. By championing deep cross-team collaboration, you will seamlessly connect ML capabilities with our Autonomy and Infrastructure organizations, breaking down silos to ensure smooth execution. Operating at the highest levels of technical maturity, you will solve the hardest technical challenges that few others can, setting the standard for state-of-the-art model distillation, advanced alignment techniques, automated agentic reasoning workflows, and the ultra-efficient deployment of billion-parameter models.

What You'll Do:

• Set the Technical Direction for Data Mining: Define the multi-year ML roadmap for our multimodal data-mining platform (Omnitag). Turn complex, ambiguous problems into clear objectives, balancing long-term foundational investments (e.g., agentic pipelines) with short-term tactical goals to drive maximum impact.
• Build the Data Flywheel: Own the full lifecycle of our ML data pipelines to close the active-learning loop. You will align massive multimodal teacher models and distill them into efficient student models for exabyte-scale search, accelerating model error diagnosis and reducing compute costs for automated curation.
• Architect Massive-Scale Systems & Tackle the Hardest Challenges: Design the architecture for applying billion-parameter models to real-world AV logs. As the lead implementer for our toughest technical bottlenecks, you will push the boundaries of parameter-efficient fine-tuning, optimize inference latency for large-scale retrieval, and build RL-driven reasoning systems—all while proactively managing system reliability and cloud footprint.
• Champion Cross-Functional Initiatives: Act as the technical bridge between Data Mining, Autonomy, and Infrastructure. Armed with a strong product sense, you will deeply understand the edge cases the autonomy stack needs to solve next and coordinate phased rollouts to seamlessly integrate new discovery capabilities into the core AV development loop.
• Lead and Grow a High-Performing Pod: Serve as a hands-on player-coach, directly guiding up to 3 engineers. Delegate ownership, coach your team to deliver state-of-the-art implementations, and advocate for tooling and practices that elevate developer productivity and reinforce Motional's culture of engineering excellence.

What We're Looking For (Must-Haves):

• BS in Computer Science, Machine Learning, or a related field, or equivalent professional experience.
• 12+ years of hands-on machine learning engineering experience, with a proven track record of owning the end-to-end development cycle of enterprise-scale ML systems, from experiment design through infrastructure, testing, rollout, monitoring, and iteration.
• Exceptional technical maturity, with a track record of building and shipping impact that is innovative, broad in reach (affecting many teams or orgs), and pays future dividends for the company.
• Proven ability to solve technical problems that few others can, the person others seek out for complex ML infrastructure, deep learning architectures, and production optimization.
• Deep experience training large-scale models from the ground up, including distributed training across multi-node GPU clusters, data pipeline and curation strategy, scaling and evaluation methodology, and the practical debugging of large training runs.
• Demonstrated ownership of a technical roadmap for a team or problem area, including balancing competing objectives (impact, quality, engineering time, compute cost) and making deliberate tradeoffs between long-term foundational investment and short-term tactical goals.
• Ability to create clarity from ambiguous, complex problems, decomposing them concisely, defining objectives, and establishing measures of progress for yourself and your team.
• Deep, generalist ML expertise spanning massive-scale model training, hardware-optimized deployment, and production ML orchestration in cloud environments (AWS, GCP, or Azure).
• Experience leading cross-team initiatives, including coordinating phased rollouts and migrations across dependent teams, and mentoring engineers to raise the bar around you.
• Strong oral and written communication skills. You can clearly explain complex technical problems using data and analysis, and communicate proposals in a way that earns buy-in and creates alignment.

Bonus Points (Nice-to-Haves):

• MS/PhD in Computer Science, Machine Learning, or a related field.
• Experience as a Tech Lead Manager (TLM) or in direct people management, coaching and growing small teams of 2–3 engineers.
• Background in autonomous driving, robotics, or complex real-time decision-making systems.
• Experience with massive-scale ML data mining, multimodal foundation models, and embodied AI.
• Hands-on experience pretraining or post-training a foundation model, data curation at scale, pretraining or continued-pretraining recipes, SFT/RLHF or other alignment methods, and evaluation harness design.
• Experience with model compression and adaptation at scale, knowledge distillation (teacher/student), parameter-efficient fine-tuning, or quantization for high-throughput inference.
• Experience with large-scale retrieval or search systems, or RL for reasoning and decision-making.
• Deep knowledge of enterprise model serving (TF Serving, Triton, TorchServe) and modern MLOps platforms.
• Experience building agentic pipelines or LLM-driven reasoning and tool-use workflows.
• A strong portfolio of publications, patents, or significant open-source contributions that demonstrates influence on the broader engineering community.