Technical Lead, Autonomy Evaluation & Metrics

ATOMS Careers page; San Francisco (CA); On-site; USD 185,000 - 242,000; Full time

Benefits offered by this job

Medical insurance

Dental & Vision

401(k)

Equity

Paid time off

Parental leave

Commuter benefits

Team lunch

Job summary

Atoms in San Francisco seeks a Technical Lead for Autonomy Evaluation to own the metrics our autonomy releases are judged on and the log replay and evaluation platform that computes them. You’ll drive how replay, scoring, metrics, and test coverage come together for on-road vehicles across software, sensor configurations, and platforms.

You will shepherd evaluation from recorded logs to scored regression suites, enabling decisions before releases reach a vehicle.

Qualifications

• Bachelor's degree with 8+ years of relevant experience.
• Experience designing safety or behavior metrics for autonomous systems.
• Hands-on experience building log replay at scale and scoring against ground truth.

Responsibilities

• Own safety and behavior metrics used to judge a release, including go/no-go criteria.
• Own the log replay and scoring pipeline for perception, localization, and planning.
• Lead regression benchmarking and dataset design across software versions.
• Develop data mining approaches to capture rare but important events in logs.
• Ensure deterministic, reproducible evaluation at scale with traceability to software versions.
• Define engineering practices and mentor new engineers in the function.

Skills

Python

C++ reading

Statistics

Leadership

Autonomous systems knowledge

Education

Bachelor's degree in Computer Science, Robotics, Electrical Engineering, Statistics

Tools

Robot middleware

Logging & replay tooling

GPU clusters

Job description

Atoms in San Francisco seeks a Technical Lead for Autonomy Evaluation to own the metrics our autonomy releases are judged on and the log replay and evaluation platform that computes them. You’ll drive how replay, scoring, metrics, and test coverage come together for on-road vehicles across software, sensor configurations, and platforms.

You will shepherd evaluation from recorded logs to scored regression suites, enabling decisions before releases reach a vehicle.