Design and architect the perception stack for autonomous construction machinery, combining sensor data and machine learning techniques to enable teleoperation, safety, and autonomy-assist systems.
Key Highlights
Technical Skills Required
Benefits & Perks
Job Description
Founding Perception Engineer — Early-Stage Robotics Startup
San Francisco, CA (onsite)
About Us
We’re a tiny, high-velocity team building autonomous capabilities for real-world construction machinery. We retrofit heavy equipment with a plug-and-play autonomy kit that hooks directly into machine controls and layers on a full sensor suite: stereo cameras, LiDAR, GNSS, and IMUs. A low-latency remote-operation network allows operators to control machines from anywhere — with autonomy layered in piece by piece.
In our first few months, we went from rough prototype to live site deployments, with real customers, real revenue, and a growing pipeline of pilot projects. We’re backed by top-tier early-stage investors and a group of technical founder-angels who believe in building hard things in the real world.
We’re led by repeat founders with backgrounds in robotics, large-scale systems, and heavy-industry operations — people who’ve shipped real hardware and built companies from zero.
Why Join Us
- You want to work on something foundational: rebuilding the physical world. Construction is slow, expensive, and stuck — and we’re fixing it the only way you can: in the field, with machines that weigh more than your apartment.
- You want to build robots in the dirt, not in slide decks. We iterate on live job sites, often from an RV or shipping container, surrounded by dust, noise, and real stakes. When your code moves 20 tons of steel, you feel it in your chest.
- You want to join a tiny founding team where every hire raises the bar — where shipping is the culture, and results beat polish every time.
About You
- You’re a hacker at heart, obsessed with fast iteration and ruthless simplicity. You know when to use geometry and when (and only when) to reach for deep learning.
- You’ve built things from scratch — a startup, a project, a prototype — and you operate with extreme personal agency.
- You work hard, stay calm in chaos, and keep a sense of humor when the excavator decides to reboot itself at 2am.
What You’ll Own
As our Founding Perception Engineer, you’ll architect the core perception stack that turns raw sensor data into spatial awareness for teleoperation, safety, and autonomy-assist systems on heavy machinery.
Architect the Perception Pipeline
- Design and evaluate 3D perception systems combining LiDAR, cameras, radar, and GNSS/IMU data for rugged, constantly changing outdoor environments.
Start Simple, Scale When Needed
- Build reliable classical baselines — clustering, tracking, rule-based filters — and evolve to ML/DL models only as data and use-cases demand.
Scene Understanding & Visualization
- Implement real-time detection, classification, and tracking to give remote operators superhuman situational awareness.
Sensor Fusion & Calibration
- Deliver robust calibration and fusion across modalities, handling dust, vibration, occlusions, and every other real-world headache.
Data & Evaluation Tooling
- Build scrappy but solid data-collection, labeling, and evaluation loops to drive rapid iteration.
Performance & Reliability
- Optimize for real-time execution on embedded compute (e.g., Jetson-class hardware) with low latency and high resilience.
Cross-Functional Integration
- Work hand-in-hand with controls, teleoperation, and safety systems to ship a cohesive human-in-the-loop autonomy stack.
Compensation
- $250–380k base salary
- 1–3% equity
- Full medical coverage, visa/relocation support, hands-on workshop resources, high-spec hardware, and all the gear you need.