Mbodi AI is hiring engineers to teach industrial robots through language

Mbodi AI, a New York robotics startup founded in 2024, is hiring three engineers to build robots that learn tasks through natural-language instruction.

The six-person company describes its product as an embodied AI platform for industrial robots. Mbodi aims to let workers teach machines new skills by speaking to them, then run those skills in production within minutes. The approach puts generative AI inside physical systems, where software must account for movement, tools, surroundings and safety.

Mbodi AI is developing its platform for companies that use industrial automation but need more flexibility than fixed robot programming can provide. Traditional systems often require specialists to define each task, adjust a robot’s motions and update the setup when conditions change. Mbodi’s system targets that training process with natural-language commands and learned behavior.

The company says its platform can help robots acquire new skills through interaction. A worker could describe a task, guide the robot through the required actions, and ask the system to repeat the process. The robot would need to connect language with perception and physical movement before it could perform the task on a factory floor.

That combination creates a demanding engineering problem. A language model can interpret an instruction, but a robot must also identify objects, control its joints, respond to unexpected conditions and complete the task within operational limits. Mbodi’s hires will work across those layers as the company develops an AI system for real industrial environments.

Top investors back Mbodi AI, and the company lists ABB among its global industrial partners. The supplied company information does not disclose a funding total. Its market position centers on the space between generative AI and industrial automation, where manufacturers want robots that can handle more varied work without rewriting the entire control system.

Mbodi joined Y Combinator’s P25 batch. The company operates from New York City and lists founders Xavier Tianhao Chi and Sebastian Peralta.

The open roles include:

• Senior Software Engineer, New York, with a salary range of $150,000 to $250,000 and equity of 0.25% to 2%. The role requires three or more years of experience.
• Founding Research Engineer, New York, with a salary range of $150,000 to $250,000 and equity of 0.25% to 2%. The role requires three or more years of experience.
• Senior Robotics Engineer, New York, with a salary range of $100,000 to $250,000 and equity of 0.5% to 2%. The role requires three or more years of experience.

The software and robotics positions point to a team building both the learning system and the machines that use it. The research role suggests work on the models and methods that connect instructions with robot behavior. The robotics role covers the physical layer, where a learned action must become a repeatable motion.

Mbodi’s pitch rests on a shift in how factories program machines. If its approach works in production, operators could train robots through demonstrations and language instead of relying on a specialist for each new task. That could help manufacturers use automation in work that changes too often for conventional programming.

The company is hiring at an early stage, with a small team, a Y Combinator affiliation and an industrial partnership with ABB. Candidates can review open positions and company information at the Mbodi AI website.