Senior Lead, Autonomy VLM

Rivian•Palo Alto, CA

About The Position

Vision-Language Models (VLMs) are a foundational pillar of our Autonomy stack. In this Tech Lead role, you will drive and deliver the overarching VLM strategy, which includes training and shipping VLM models, extending to multi-modalities, enabling new use cases, among others. In this role, you will also be responsible to architect VLM-driven solutions to solve some of autonomy's hardest challenges, including automated data mining, handling long-tail distributions, rare edge-case detection, and scene anomaly reasoning. You will also drive our large-scale training data acquisition strategy for VLM-related model training, closely collaborating with our teams and partners. You will also own the whole end-to-end lifecycle of VLM model delivery: data acquisition, metrics definition, benchmarking, model performance optimization, deployment, feedback loop. Collaborating broadly across the Autonomy org, you will serve as the champion for VLM models and data mining capabilities, as well as represent these efforts in our interactions with other teams

Requirements

BS, MS, or PhD in Computer Science, Robotics, Electrical Engineering, or a highly related quantitative field.
5+ years of professional experience scaling ML solutions, with a strong focus on the following:
VLM model training: Hands-on experience training or fine-tuning VLMs using modern parameter-efficient techniques (LoRA, QLoRA) and RL alignment.
Large-scale data mining: Proven track record developing VLM/LLM-related techniques for data mining, long-tail distributions, rare cases, safety-critical events.
Zero/few-shot capabilities: Experience with open-vocabulary, zero-shot, or few-shot classification models, particularly in long-tail scenarios.
Training data strategy: Experience with driving training data acquisition strategy to train VLM-related models, defining data annotation guidelines, partnering effectively with in-house and external 3P annotation vendors.
System engineering: Strong proficiency in Python alongside a solid understanding of modern Perception pipelines, benchmarking tools, and infrastructure.
Execution: Demonstrated ability to root-cause complex issues across a distributed, cross-functional stack in a fast-paced environment.

Nice To Haves

Experience applying VLMs within the Autonomous Vehicle domain.
Experience with Auto Prompt Optimization (APO) and automated prompt engineering techniques.
Experience with spatial grounding in 2D and/or 3D.
Experience extending foundational models to extra modalities (e.g., LiDAR, Radar, IMU, ego-motion).
Experience utilizing VLMs or Foundation Models for complex behavior reasoning and planning.
Experience with onboard edge deployment, cloud inference architectures, and balancing compute/efficiency trade-offs.
Experience with quantization techniques (PTQ, QAT) and high-performance inference engines like TensorRT

Responsibilities

Drive and deliver the VLM strategy: Own the holistic roadmap of the VLM strategy, including training and delivering VLM models, deployment, alignment, and ensuring a unified vision across the Autonomy org.
Accelerate data mining: Design and deliver VLM-related models and strategies that power automated data mining, long-tail distributions, rare/edge case detection, and anomaly detection at scale.
Drive and deliver the data acquisition strategy: Architect the strategy for large-scale training data acquisition to train the VLM models and improve their performance, establishing workflows with in-house and 3rd-party annotation vendors.
Iterate and optimize performance: Establish rigorous evaluation and monitoring benchmarks. Identify and root-cause top-tier system anomalies, prioritizing high-impact optimizations to continuously push the needle on performance.
Cross-functional collaboration: Partner closely with core Autonomy teams (Perception, Planning, Calibration, Systems, etc) to translate vehicle feature requirements into concrete ML deliverables.
Influence trade-offs & requirements: Define system requirements and guide cross-functional efforts through technical trade-off decisions