Senior, Software Engineer - AutoTagging

About the Company

At Torc, we have always believed that autonomous vehicle technology will transform how we travel, move freight, and do business. A leader in autonomous driving since 2007, Torc has spent over a decade commercializing our solutions with experienced partners.   Now a part of the Daimler family , we are focused solely on developing software for automated trucks to transform how the world moves freight.  Join us and catapult your career with the company that helped pioneer autonomous technology, and the first AV software company with the vision to partner directly with a truck manufacturer.

Meet the Team

The Auto Tagger team is the engine behind our data flywheel, responsible for translating petabytes of raw, multi-modal vehicle data into a highly curated library of critical driving scenarios. By mining driving logs for long-tail events, we provide the foundational data required for safe autonomous trucking. Leveraging Pegasus logical layers, this team structures and catalogs findings into an observations database that directly accelerates development across autonomous perception, sensor fusion, and generative simulation testing.

What You'll Do

Design and implement event-detection algorithms over multi-sensor time-series data (ego kinematics, perception object tracks, lane and map geometry) using temporal logic, state machines, signal processing, and geometric reasoning in road-relative frames.

Combine rule-based logic with learned components (classifiers, embeddings, Vision-Language Models) where heuristics hit their ceiling, and make the call on which approach fits each scenario.

Build evaluation for the taggers you write: ground-truth sets, per-scenario precision and recall, failure analysis, and regression tests that catch quality drift.

Turn ambiguous stakeholder requests (“find interesting merges”) from perception, simulation, and systems teams into precise, testable scenario definitions mapped to Pegasus layers.

Own your taggers end to end into production: Python on Databricks/Spark, CI via GitHub Actions, and monitoring of both tag quality and job health.

Optimize taggers for scale and cost across petabytes of logs, favoring vectorized, distributed implementations over per-frame processing.

What You'll Need to Succeed

Bachelor’s Degree in Computer Science, Robotics, Electrical Engineering or related technical field plus demonstrates competences and technical proficiencies typically acquired through 6+ years of experience OR Master’s Degree in Computer Science, Robotics, Electrical Engineering or related technical field plus demonstrates competences and technical proficiencies typically acquired through 3+ years of experience.

Demonstrated algorithm design on real-world sensor, time-series, or geospatial data, in areas such as perception, tracking, state estimation, prediction, planning, signal processing, or applied ML.

Strong Python skills, including NumPy, pandas, and SciPy, plus working proficiency in SQL.

Experience defining metrics and evaluating algorithm quality against labeled or ground-truth data.

Hands-on exposure to production data systems: distributed compute (Spark, Ray, or Databricks), a cloud platform (AWS preferred), and CI/CD.

Strong communication skills to work across ML, perception, and simulation teams.

Bonus Points!

Autonomous vehicle or robotics data experience: ROS bags/MCAP, HD maps, Frenet or lane-relative coordinates.

Familiarity with scenario description standards such as Pegasus layers or OpenSCENARIO.

Experience with VLMs, zero-shot classification, auto-labeling pipelines, or embedding/vector search for scenario extraction.

Experience with GitHub Actions, infrastructure-as-code (Terraform or CloudFormation), and observability tooling (Datadog, Grafana, CloudWatch).

Experience with model serving frameworks such as vLLM or SGLang.