Full-Stack Robotics Engineer (Forward Deployed)

About the role

Show up at a farm, a cut floor, a cable plant, or a mine, find the job nobody has been able to automate, then design the robot that does it, build it with your own hands, and put it into production.

Synphony is the deployment layer for physical AI. Already having launched from Y Combinator with multi-million dollar deals, we take frontier models—VLAs, foundation policies, and agents—and make them work inside the messy physical industries that run the real economy. Not in a lab, not in a video. In heat, dust, vibration, bad lighting, legacy PLCs, and in front of people who have done the job by hand for thirty years and do not care about your architecture.

Responsibilities

• Scope manual tasks on the customer’s floor with their engineers.
• Design and build the machine (mechanism, kinematics, frame, end effector, tooling, fixturing) in our warehouse—machined, printed, and wired by you.
• Instrument the cell, collect data, train and fine-tune policies (vision-language-action models, imitation learning, RL, residual policies).
• Solve sim-to-real challenges, wire the system into customer infrastructure (PLCs, MES, ERP, operator dashboards).
• Commission the system on a live line, train operators, and ensure remote debug capability.
• Own uptime post-deployment and iterate across customers/industries.
• Track and reproduce frontier advancements in physical AI (compute/hardware provided).

Requirements

The bar—you’ve likely done several of these:

• Designed and built a working machine (not just a simulation) and deployed it in an uncontrolled environment.
• Machined, printed, or wired your own parts to avoid delays.
• Trained and deployed a real learned policy on hardware (not just in sim) to meet customer rate requirements.
• Hands-on with imitation learning, RL, or VLA fine-tuning (reward shaping, offline RL, DAgger, LoRA, action tokenization, edge inference).
• Worked with contact-rich tasks (force control, compliance, calibration, deformable/irregular objects).
• Integrated systems into equipment you didn’t own (PLCs, MES, ERP) with no API/docs/help.
• Stood up ROS/vendor SDKs, cloud/GPU infrastructure, and customer-facing dashboards.
• Owned a machine through acceptance testing, hit cycle-time targets, and maintained post-handover uptime.
• Translated a $4M/year labor/turnover problem into a working solution for skeptical operators.

What isn’t optional:

• Raw building ability and autonomous creative problem-solving.
• Professional fluency in English (you’ll explain solutions in loud, high-stakes environments).
• Never saying “that’s not my area.”

Challenges

• Heavy travel to hot, loud, dusty sites with poor conditions.
• Robots break in unpredictable ways; policies may regress due to undocumented variables (e.g., faster conveyors).
• Data is often messy; skepticism from experienced operators until you prove results.

Why it's worth it

• Own the entire physical-AI stack—from raw sensor data to deployed systems—across multiple industries.
• Work directly with customers as the face of the company, compounding real-world data into better models.
• Tackle the “dirty jobs” in food, manufacturing, agriculture, mining, and logistics that others avoid.