Operations Lead - Physical AI Data (APAC)

Because before AI can be trusted, its training data must be. Our work sits at the intersection of egocentric video understanding, embodied AI, robotics perception, and voice-driven interaction. We move fast, obsess over data quality, and ship at scale.

This role is central to how we build the datasets that train the next generation of Physical AI systems. If you have managed large-scale operations with multiple sites and stakeholders — and want to do it at the frontier of AI — this is your role. The Role We’re hiring an Operations Lead to own Physical AI data operations across APAC, based in Bangalore.

You’ll stand up and scale data collection across multiple sites and multiple countries — recruiting and managing crews, securing site access, running the full capture-to-upload pipeline, and hitting throughput, yield, and cost targets. The mandate is to scale APAC collection toward 1M hours per month. This is the person who can sit across from a factory owner to negotiate floor access in the morning and stand next to a line worker fixing a rig at 10pm.

You think in systems and SOPs. You’ve run large, distributed, physical operations before, and you know the difference between 40% and 95% usable output is process, not luck. What You’ll Own Field collection at scale Stand up and run egocentric video collection across APAC sites — manufacturing floors, agricultural operations, warehouses, residential settings — across multiple countries and languages.

Secure and manage site access: negotiate with factory owners, site managers, and local partners, and build the relationships that let us collect inside live operating environments. Recruit, onboard, train, and manage field crews and site leads. Build the playbook that takes a new site from zero to producing usable data in days, not months.

Throughput, yield, and cost Own the metrics that matter: usable hours collected, discard rate, cost per usable hour, and time-to-first-data on every new site. Drive usable yield up by pulling quality control into the capture loop — in-language feedback to contributors before a session ends, not flagged in post. Hit collection targets across concurrent programs and geographies, and forecast capacity as we scale.

SOPs and systems Build collection SOPs at the task-environment level — what a warehouse capture needs is not what a manufacturing capture needs. Codify each so it’s repeatable across sites and contributors. Own hardware and data logistics end to end: rig allocation and rotation, device health, charging and thermal management over multi-hour shifts, physical drive rotation and secure offload, and the upload pipeline and chain of custody from capture to delivery.

Product and cross-functional Carry a strong voice in product development — both hardware (rig design, ergonomics, camera configuration, capture devices) and software (capture, QC, upload, and ops tooling). You’re the one running it in the field, so the roadmap is shaped by what you learn there. Partner with our contributor network and recruiting engine to keep sites supplied with verified, qualified contributors.

What We’re Looking For Must-have 6 years running large-scale physical operations, with proven ownership of distributed, multi-site execution. Backgrounds we like to see: e-commerce or supply-chain logistics at scale, dark stores / quick-commerce, fleet operations (ride-hailing, delivery, logistics), or multi-factory / multi-site manufacturing operations. We’re open to less experience for someone who has demonstrably built systems from scratch.

Track record of building operations from the ground up - SOPs, hiring, training, and the systems that hold quality and throughput as you add headcount and sites. Comfortable across the full stack of stakeholders. You can earn the trust of a factory owner and a floor worker in the same visit, and get both to do what the operation needs.

Systems thinker, obsessive about process and measurement. You instrument everything and use the numbers to catch problems before they hit output. Willing to travel across APAC and spend real time on the ground.

This is not a desk role. Fluent operating in India and comfortable working across languages and cultures in the field. Strong plus Experience managing a blue-collar or gig workforce at scale — recruiting, incentives, retention, and quality accountability.

Experience running hardware or device fleets in the field — allocation, maintenance, repair, and logistics. Exposure to data collection or any QA-heavy operation where output quality is measured and enforced at scale. Cross-border operations experience across APAC manufacturing or logistics hubs.

Who You Are You build a structure where none exists and leave behind a system that runs without you. You’re a relentless problem-solver — equally at home fixing a field blocker and presenting an ops review to leadership. You measure everything, and you own outcomes, not activity.

You move fast and you care about quality, because the models are only as good as the data going into them. Why Humyn Labs Own one of the hardest and most consequential operational challenges in AI — the field infrastructure that trains physical-world intelligence, in the markets where it deploys first. A genuine zero-to-one mandate: scale APAC collection toward 1M hours per month, with the autonomy to build the system that gets there.

Operational depth is our moat, which makes this role core to the company, not a support function. Tight-knit, high-calibre team with deep roots in AI research, robotics, and large-scale data operations. Competitive compensation and equity in a company building at the frontier.

Bengaluru office — collaborative, in-person, high-energy environment. A culture that rewards ownership, speed, and obsession over outcomes.