Robotics Data Pipeline & Annotation Lead

Genesis; San Carlos (CA); On-site; USD 180,000 - 240,000; Full time

Job summary

Genesis, a frontier AI robotics company, seeks an experienced data and ML pipeline lead to own the data engine behind Eno's training data. You will design and operate the end-to-end workflow turning raw robot demonstrations into labeled, training-ready data, with automation to scale beyond headcount.

You will define datasets and ontology with the model team, apply vision-language models for annotation and data synthesis, and drive quality metrics like inter-annotator agreement and throughput.

Qualifications

• Four or more years in data or ML pipelines, including time leading the work.
• Strong Python (Pandas, NumPy, PyTorch) and SQL. You write the automation that shrinks the pipeline.
• ML literacy: you understand training vs test, precision/recall, and design ontology that helps the model.
• Hands-on technical leadership: you can run a labeling operation and stay a hands-on contributor.
• Comfortable with ambiguity and speed in a research-paced environment.

Responsibilities

• Run the data engine: manage raw trajectory and video to training-ready datasets with validation steps.
• Own datasets and ontology: decide what gets annotated and how, designing the ontology with the model team.
• Automate with models: use vision-language models for automated annotation and data synthesis to scale the pipeline.
• Run the annotation operation: scale labeling, internal and vendor, to a clear quality bar and delivery schedule.
• Close the loop: turn real-robot eval failures into targeted collection and annotation jobs.
• Own the metrics: track inter-annotator agreement, label error rate, and throughput per annotator-hour.

Skills

Python

SQL

Pandas

NumPy

PyTorch

ML Pipelines

Tools

Airflow

Vision-Language Models

Job description

Genesis, a frontier AI robotics company, seeks an experienced data and ML pipeline lead to own the data engine behind Eno's training data. You will design and operate the end-to-end workflow turning raw robot demonstrations into labeled, training-ready data, with automation to scale beyond headcount.

You will define datasets and ontology with the model team, apply vision-language models for annotation and data synthesis, and drive quality metrics like inter-annotator agreement and throughput.