About Human Archive
Human Archive is a robotics data lab founded by Stanford and UC Berkeley dropouts. We work alongside frontier robotics labs and foundation model research groups to collect large-scale, real-world, annotated multimodal datasets of humans performing everyday tasks across household and industrial environments.
We are lean, technical, and operate at extreme speed, taking on unglamorous and conventionally impossible problems that directly unlock step-function gains in model capability.
The deployment of capable humanoids at scale will permanently redefine human labor. Undesirable physical work will disappear, and human effort will shift toward a new era of abundant creativity. This shift is inevitable, and we are building the infrastructure to accelerate it.
We are assembling the best team to solve the hardest problems in embodied intelligence. You will own meaningful systems from day one and see your work directly impact model capabilities. This is a once-in-a-generation inflection point. If you want to leave your dent on humanity and reshape physical labor markets forever, join us!
The Opportunity
The head of engineering will own the entire technical execution of a body-worn sensing platform — from architecture through shipping — while leading a distributed team of specialists across San Francisco and India.
The head of engineering will make real design and integration decisions every week across all four core domains: power systems, sensor synchronization, firmware and embedded Linux, and mechanical structure.
RELEVANT EXPERIENCES MUST BE BOLDED IN RESUME
Power systems: You have personally designed a multi-rail battery-powered architecture — not reviewed someone else's. You can size a protection circuit, choose a regulator topology for a given load profile, and explain the inrush and brownout behavior you encountered on first power-on. You can read a schematic and find a problem.
Sensor synchronization: You have solved the problem of aligning data from multiple sensors with mismatched sample rates and variable bus latencies. You know what a time-stamping strategy looks like at the firmware level, and you know what misalignment actually looks like in a captured dataset — not in theory, in your own data.
Firmware and embedded Linux: You have personally brought up a SoM or SBC platform under real workload. You can read and follow firmware written by someone else well enough to identify a timing bug, interrupt priority issue, or buffer overrun using a logic analyzer or serial log. You do not need to write production firmware, but you must be able to debug it.
Mechanical judgment: You will not own CAD in this role, but you must review, critique, and push back on mechanical designs with engineering justification. You understand load paths, tolerance stack-up, strain relief, and what "designed for the lab" looks like vs. designed for a person wearing it for eight hours.
We’re archiving the physical world for embodied intelligence by collecting and labeling aligned multimodal data. To build dexterous and perceptive robots that generalize robustly, we need massive amounts of real-world data across multiple modalities and environments.
We have thought deeply about the fine line between biomimicry and its application to humanoid systems. Based on this research, we design and deploy custom hardware across residential and manufacturing settings. We then post-process the resulting data through internal QA, anonymization, and annotation pipelines to deliver diverse, high-fidelity datasets at scale to frontier labs developing robotics foundation models and general-purpose robotics companies.
We believe we are at a historic inflection point, with a unique opportunity to leave a dent on humanity and reshape physical labor markets forever. That's why our team dropped out of Stanford and Berkeley and moved to Asia to collect the world’s largest annotated multimodal dataset.
Salary
$140,000 - $210,000
Equity
0.25% - 2%
Location
San Francisco, CA, US / IN
Experience
6+ years
Last stage
Seed
Investors
No applications, no recruiter spam. Just the intro.
A few questions to make sure this role is the right shape for you. Two minutes.
I write the intro, send it to the founder, and handle the back-and-forth.
If they’re a yes, I book the chat. You show up — that’s the whole job-hunt.