Amazon · amazon.jobs · checked today
Senior Robotics Data Engineer, Frontier AI & Robotics (FAR)
Are you interested in solving the most critical operations problem behind cutting-edge robotics development? Interested in fast dynamic environment? Amazon's Frontier AI & Robotics (FAR) team is building advanced robotic systems through frontier foundation models, and high-quality data collection is a core input to that research.
Skills, with evidence
- Data quality
Partner with researchers to translate research objectives into reliable collection workflows, improve data quality, and shorten policy training iteration cycles.
must have - Machine learning
Experience supporting ML/AI data collection, annotation, quality assurance, or model evaluation workflows.
must have · not practised here - Data modelling
Maintain version control and ensure collectors adopt changes consistently.
must have - AWS
Amazon's Frontier AI & Robotics (FAR) team is building advanced robotic systems through frontier foundation models, and high-quality data collection is a core input to that research.
not practised here - Governance & security
Assess operational and technical risks, including equipment readiness, unsafe conditions, data loss, and quality degradation.
not practised here - Failure handling
- Partner with researchers to develop and refine SOPs, translating task objectives, failure modes, and evaluation feedback into clear collection instructions and acceptance criteria.
- Streaming
Lead operational improvement projects that reduce collection failures and iteration delays.
- Dashboards & BI
Use dashboards and root-cause analysis to identify recurring issues and verify that corrective actions improve results.
not practised here
Your plan
- The SQL screen: correct, then fast≈ 2 h
Data quality
- Late arriving event auditIntermediate
- Order anomaly auditIntermediate
- Premium and free listening by countryIntermediate
- Interview audit union capstoneAdvanced
- Airbnb: converting on a day with no published rateAdvanced
- Data modelling: the round most people fail≈ 3 h
Data modelling
- Addresses that stay true to the pastIntermediate
- Seat holds and the release-night raceIntermediate
- Subscription warehouse grainIntermediate
- Campaign efficiencyIntermediate
- Catalogue: products, variants and sellersIntermediate
- Pipeline design: safe to run twice≈ 4 h
Failure handling · Streaming
- The source will not let youIntermediate
- Parcel tracking pipelineIntermediate
- Is this change safe?Intermediate
- Five minutes behind the sourceIntermediate
- Changing a pipeline that’s already runningIntermediate
- Say it out loud≈ 1 h
15 drills · Intermediate + Advanced≈ 10 hours
Not covered by the plan: Machine learning, AWS, Governance & security, Dashboards & BI.
Readiness
Counted from drills you have completed anywhere on D8LooP.
leaves in 4 dremoved the moment Amazon closes it