Robot.com CEO Felipe Chavez says robots should replace boring labor, not people’s careers

Felipe Chavez, co-founder and CEO of Robot.com, is making a very specific promise: robots should take the physical, repetitive labor that people do not want to do. During a Business Insider visit to Robot.com’s San Francisco headquarters, Chavez described his “idyllic yogurt commercial” style vision, where humans pursue meaningful lives because automation handles the superfluous and mundane. It is an optimistic framing for a hard-nosed operational bet, and it is already showing up in how Robot.com builds and sells robots.

Chavez says Robot.com has more than 500 working robots deployed, with a majority handling delivery. The company also says it has completed more than 2.5 million tasks. For investors, operators, and boards watching robotics and labor markets, the interesting part is not just the scale. It is the thesis behind the scale: the path to adoption is not “general-purpose humanoid robots everywhere,” but task-by-task automation that removes high-turnover chores first.

Robot.com’s starting point is delivery. Chavez said about 400 of its robots are still delivery robots deployed across college campuses. But the company is not staying in that box. Chavez described expansion into warehouses, food service, and advertising, including robots acting like roving billboards. In other words, Robot.com is treating “robot work” as a product line with different form factors, rather than as one machine that solves all problems.

That product thinking comes from an operational lesson Chavez says he learned the hard way while building his first startups. Before launching Kiwibot in 2017, which later became Robot.com, he ran a grocery-delivery company in Colombia that was bootstrapped, sometimes requiring him to complete deliveries himself. He said he realized the manual work could be exhausting and boring. Fast forward to today, and Robot.com is trying to convert that human experience into a robotics strategy: automate the repetitive and mundane tasks that often correlate with turnover.

The company’s most telling example is how automating one job can create a new job. Chavez described a scenario from its food-delivery business: as robot-delivery scaled, restaurant workers found themselves with an added chore of walking outside to place to-go containers inside the robots. That created friction. In practice, he said customers had to retrain workers or even hire extra people, which is a classic failure mode in operations automation. The fix, in Chavez’s telling, is not to declare victory after first automation. It is to keep decomposing the workflow until you find the next manipulable bottleneck.

That is why Robot.com talks about analyzing labor by breaking it into specific tasks. Chavez said the company follows a rule of thumb for its robotics pipeline: “If you can do it with two fingers, very likely we will be able to do it.” This is a deliberate contrast with the broader Silicon Valley humanoid push. Companies like Figure AI and Tesla are building bipedal robots with five fingers, which means they have to solve the herculean problems of full dexterity and manipulation before wide deployment in homes, warehouses, factories, and other physical environments. Chavez’s near-term bet is more grounded: sorting, moving, handing items off, and other less glamorous service work that is widespread and operationally repeatable.

Chavez also anchored the human side of the story in incentives. He said Robot.com’s current customers, about 20, are thinking less about replacing workers and more about improving workforce satisfaction and reducing turnover. “The people that are working right now are going to feel better,” he said. “They’re going to maybe not do that repetitive tasks and focus on customer-centric experience and not resign after seven months.” That is a measurable business concern, even if it is framed with a kind of solarpunk optimism: keep workers in the loop, reduce the churny parts of service work, and let humans do the parts that require judgment, interaction, and guidance.

Of course, the labor math is real. The article cites estimates from the Bureau of Labor Statistics: there were about 3.8 million fast-food and counter workers, 3 million laborers and freight stock and material movers, and 2.8 million stockers and order fillers in 2024. Even if humanoids never arrive in the way hype suggests, there is still a massive universe of “simple jobs” that could be partially automated. Other robotics companies have made similar “humans upskilled, not replaced” claims, but they still face the same economic reality: large labor forces exist because the work is widespread, and widespread work is where automation either delivers real relief or triggers real backlash.

Chavez’s argument tries to split the difference. He said humans will remain essential in a robotic future, pointing to examples like humans maintaining and guiding Waymo robotaxis at remote centers or training humanoids through teleoperation. In Robot.com’s ad imagery, humans peacefully co-exist with robots, potentially even receiving a coffee along their commute. The strategic takeaway is clear: Robot.com is positioning itself to sell robotics by focusing on workflow replacement that can be justified today, then expanding from delivery into warehouses, food service, and advertising. For executives making bets in robotics, logistics, and labor-adjacent tech, the question is no longer whether automation can happen. It is whether the first automations are chosen well enough to reduce turnover rather than simply shifting work around.