Perception Engineer, Sensor Fusion

About the role

What you’ll do

• Design and ship late and mid-level fusion of lidar, radar and camera detections into stable multi-object tracks with calibrated covariances.
• Build state estimators and data association (Kalman variants, JPDA, learned affinity) that survive occlusion, split-and-merge clusters and radar multipath.
• Diagnose perception failures from fleet logs, separating sensor, calibration, timing and model causes before proposing a fix.
• Define metrics for tracking quality and calibration of confidence, and wire them into replay-based regression gates.
• Work with embedded engineers to hit latency and jitter budgets for the perception pipeline on the vehicle compute.
• Review model and tracker changes from colleagues with attention to failure modes at night, in rain and in glare.

What you’ll bring

• Several years building perception or tracking systems that ran on real vehicles or robots, not only on public benchmarks.
• Working knowledge of probabilistic state estimation, including EKF or UKF design, gating, track management and consistency tests such as NEES.
• Strong C++ in performance-sensitive code, plus Python for analysis, with comfort profiling latency and memory on target hardware.
• Familiarity with point cloud processing, radar signal characteristics and camera geometry, including extrinsic and temporal calibration.
• Experience reasoning about sensor timestamps, motion compensation and clock offsets as a first-class source of error.
• A habit of measuring a change against a held-out scenario set before calling it an improvement.

Nice to have

• Experience with learned detectors on bird's-eye-view or voxel representations and fusing their outputs with classical trackers.
• Background in radar perception, including Doppler-based ego-motion and clutter rejection.
• Experience with conformal prediction or other methods for calibrated detection confidence.
• Contributions to open-source robotics middleware or perception libraries.

Work you might take on

1. 01

   Replace a nearest-neighbour association step with a learned affinity model and show fewer identity switches on dense urban replay.

2. 02

   Add radar Doppler as a constraint in the tracker so slow-moving pedestrians near stationary clutter keep a stable track.

3. 03

   Build a per-object confidence calibration report that the planning team can use to set conservative gaps around uncertain tracks.

4. 04

   Track down a rare phantom-object event in heavy rain and land a filter on the lidar returns that removes it without clipping real targets.

Compensation

$220K – $360K

The US base salary range for this role. Every offer also includes equity in Clexerl and, when it spins out, in the venture you help build. We tell you the reasoning behind your number.

Benefits

• Health, fully covered

  Medical, dental and vision for you and your dependants, premiums paid in full.

• Two kinds of equity

  Equity in Clexerl, plus a stake in the venture you help build when it spins out.

• 22 weeks parental leave

  Fully paid, for every parent, however your family arrives.

• Rest that sticks

  Flexible time off with a four-week minimum we actually check on, plus a company-wide week in December.

• Learning and compute

  A yearly budget for courses, conferences and books, and compute credits for side projects.

• 401(k) with a 4% match

  Matched from your first paycheque and vested immediately.

How we hire

1. Apply 5 minutes

   Send links to what you have built and a few lines on why this role. No cover letter needed.

2. Intro call 30 minutes

   A conversation with the hiring manager about the work, the team and what you want next.

3. Work sample 2–4 hours

   A paid take-home or a pairing session on a real-shaped problem. You pick the format.

4. Studio day One day

   Meet the team, go deep on your work and ours. Remote or in person.

5. Offer Within 5 days

   A written offer with the full range, the equity and the reasoning behind both.