Robots Are Taking Over Some Human Jobs. But Some Are Being Hired To Help Them Complete Tasks.

Robots are increasingly being used to automate work in factories, warehouses and other businesses. But many still struggle with tasks that humans can handle easily, creating an opportunity for a San Francisco startup that pairs robots with people who can operate them remotely.

Aktoria Robotics is building a network of trained human operators who can take control of robots when their automated systems fall short. Its service, called "Human as API," allows robotics companies to get assistance without hiring and managing their own remote-control teams.

The startup, backed by Y Combinator, operates across the United States, India and Mexico. It claims companies can save at least 50% compared with maintaining in-house operators, although the savings have not been independently verified.

Aktoria is betting that robots will continue needing human assistance even as they become more capable. The company was founded this year by Jaskaran Singh Walia and Shreenabh Agrawal, two researchers with backgrounds in robotics and machine learning at Carnegie Mellon University.

Aktoria's technology uses a process known as teleoperation, which allows someone to control a robot from another location. An operator can step in when a machine encounters an unfamiliar object, gets stuck or cannot complete a task using its existing software.

Instead of requiring a trained specialist to be physically present, the robot can receive assistance through an internet connection.

The company says its technology can maintain a delay of approximately 130 milliseconds between an operator in India and a robot in San Francisco, although that figure comes from its own testing. The company also claims cross-country response times as low as 40 milliseconds, according to its Y Combinator launch announcement.

Remote operation is one part of the business. Aktoria is also collecting data from human-controlled robot movements that developers can use to train artificial intelligence systems.

When an operator helps a robot complete a task, the interaction can generate recordings of the machine's movements, camera feeds and sensor information. Such datasets can help robotics developers improve how machines respond to unfamiliar situations.

Aktoria advertises quality-checked recordings of real robot operations through its official website, giving the company another potential source of revenue beyond providing human operators.

The business is emerging as North American companies continue investing in automation. Robot orders in North America rose 24.8% in the third quarter of 2026 to 10,989 units, with a combined value of $723.5 million, according to the Association for Advancing Automation.

The value of orders increased 26.1% from a year earlier, with demand growing among automotive suppliers, semiconductor manufacturers and electronics companies.

However, purchasing robots does not necessarily mean businesses can operate them without human involvement.

A September study by Anthropic estimated that existing robots are technically capable of performing about 74% of physical work tasks in the United States, but often only in controlled environments.

The researchers found that robots were currently cost-competitive with human workers for just 0.3% of work tasks, partly because of their high costs and technical limitations.

Those limitations have created demand for companies offering remote monitoring and intervention services.

Aktoria is not alone in this market. Robotics software company Formant offers remote fleet management and teleoperation tools, while FORT Robotics expanded its remote supervision capabilities through its acquisition of Mapless AI.

Aktoria is seeking to compete by combining its own operators with remote-control software and training data.

The startup has not publicly disclosed its revenue or named major customers, making it difficult to independently assess the scale of its operations or the financial benefits it promises.

Aktoria is also hiring as it develops its technology, advertising a robot learning researcher position in San Francisco with a salary range of $450,000 to $600,000, alongside engineering and operations roles in India.