T
Torc Robotics develops autonomous vehicle software for Class 8 trucks in partnership with Daimler, the largest heavy-duty truck manufacturer in North America. The stack reveals a perception-first architecture: LiDAR + multi-modal sensor fusion, PyTorch-based deep learning models, simulation infrastructure (neural rendering for perception, offline annotation pipelines), and automotive safety standards (ISO 26262, AUTOSAR). Active hiring is heavily engineering-focused, with nearly 80% of open roles in engineering and a notable adoption of ROS 2 and Bazel—infrastructure choices that suggest a shift toward modular, scalable autonomy architecture as the company moves beyond early R&D.
Tech Stack 110 technologies
Core StackAWS Terraform Pandas Python PyTorch Tableau C++ Temporal Docker Lightning Ray AWS EKS Ray Data Parquet PyArrow LanceDB Gymnasium LiDAR CI/CD Bazel Debian Dev Containers Yocto Infineon AURIX AUTOSAR IBM DOORS Jama Connect Automotive Ethernet I2C ISO 26262+79 more
AdoptingROS 2 Bazel
What Torc Robotics Is Building
◆Challenges
- Reducing functional insufficiencies
- Perception accuracy improvement
- Domain data gap
- Filling domain data gaps
- Ensuring safe autonomous driving
- Reducing technical debt
- Streamlining development workflows
- Scaling perception simulation
- Reducing calibration obstacles
- Ensuring consistent performance and safety standards
▲Active Projects
- Road & lane detection model roadmap
- Build automation tooling and pipeline development
- Neural rendering framework for perception simulation
- Simulation integration
- Offline perception annotation pipeline
- Multi-modal sensor fusion architecture
- Data ingestion and governance
- Offline perception model development
- Complex workflow automation via temporal bazel, buildstream
- Autonomous driving platform
Hiring Activity
75 roles · 60 in 30d
Department
Seniority
Notable leadership hires: Engineering Director
Company intelligence