AI Researcher - Perception

1X
United States
Workplace: OnsiteFull timeUSD 200,000 - 300,000 annuallyFunction: Research & Scientific (R&D)Experience: 5+ yearsSkills: ["Technical direction","Tradeoff analysis","Hands-on problem solving","Code review"]

Own NEO’s perception systems end to end, from first prototype through fleet-wide deployment. Build vision pipelines that recover robot geometry, human and object pose, and reconstruct 3D scenes and terrain for terrain-aware locomotion and contact-rich manipulation. Deliver depth and scene understanding, deploy pose and perception models in production, and validate improvements based on real robot behavior, not just offline metrics.

Loading

Loading job details...

Preparing the role view and application actions.

FursaFursa
1X
1X
1 day ago

AI Researcher - Perception

✓ Verified Job

Canonical indexed version, validated from employer's careers page.

Source: Company careers pageValidated by: Fursa AI
Last checked: 13 hours agoStatus: Live

Job Summary

Own NEO’s perception systems end to end, from first prototype through fleet-wide deployment. Build vision pipelines that recover robot geometry, human and object pose, and reconstruct 3D scenes and terrain for terrain-aware locomotion and contact-rich manipulation. Deliver depth and scene understanding, deploy pose and perception models in production, and validate improvements based on real robot behavior, not just offline metrics.
Location: United States
Workplace: Onsite
Employment Type: Full time
Job Function: Research & Scientific (R&D)
Seniority: Mid level

Key Responsibilities

  • •Own NEO perception from first prototype through fleet-wide deployment, measuring impact on deployed robots.
  • •Build vision systems to recover robot geometry, human and object pose, and reconstruct terrain and scenes for locomotion and manipulation.
  • •Deliver depth and 3D scene understanding and ship pose estimation models usable for large-scale data labeling and motion generation.
  • •Implement sensor fusion with correct time synchronization across cameras and inertial sensors for consistent downstream data.
  • •Run camera calibration at scale (intrinsic/extrinsic) with monitoring for drift and recoverability via field tools, and define datasets/metrics to evaluate behavior improvements.

Pay and Benefits

Salary: USD 200,000 - 300,000 annually
Equity and Bonus:Equity
Perks:Health InsuranceDentalVisionPaid LeaveParental Leave401kFsaHsaDisability InsuranceLife InsuranceEap

Key Requirements

  • •Proven track record shipping perception systems that run on physical hardware in the real world.
  • •Demonstrated ability to train, evaluate, and deploy deep learning models for detection, depth, or pose estimation in PyTorch.
  • •Deep experience with geometric computer vision and camera calibration (intrinsics/extrinsics), including kinematic calibration (hand-eye/foot-eye).
  • •Experience owning a capability end to end, from problem definition through deployment and iteration on real failures.
  • •Strong Python and modern C++ in large shared codebases, with solid testing and code review practices.
Experience:5+ yearsRobotics
Skills:Technical directionTradeoff analysisHands-on problem solvingCode review
Tech Stack:PyTorchPythonC++TransformerDeep learningGeometric computer visionMulti-view geometryNon-linear optimizationKinematic calibrationHand-eye calibrationFoot-eye calibrationIntrinsicsExtrinsicsSensor fusionVisual-inertial odometrySLAMSegmentationClassificationOpen-vocabularyFoundation backbones

Company Brief

1X
Develops humanoid robots and embodied AI systems designed to perform useful work in real-world environments. The company focuses on building general-purpose robotic labor for tasks in industries such as logistics, manufacturing, and security.
Industry: Robotics
Company Size: Medium (51 to 250 employees)
Growth: Growth Stage Startup
Headquarters: Moss, Norway
Founded: 2014
WebsiteLinkedIn