Feather Robotics is developing a modular robot platform that lets companies deploy their own physical AI applications without first building a complete robot from scratch. Founded in 2025 by Hoa Mai and Parsa Bakhtiari, the startup has already generated more than $1 million in revenue from early customers.
Feather is taking a different approach from companies such as Tesla and Figure, which are developing both robot hardware and the foundational AI systems controlling them. Instead, Feather provides developers with configurable hardware, software tools and support for AI models from providers including Nvidia, Skild and Physical Intelligence.
Feather Focuses on Modular Hardware for Developers
The company’s robot platform uses two seven-degree-of-freedom arms mounted on an adjustable torso and an omnidirectional wheeled base. Developers can modify components such as arm configurations, computing hardware, cameras and end effectors for different applications.
Feather says the system is already operating in real-world environments after about a year of field testing. The company has not disclosed its customers but said its robots are being used to cook in restaurants and clean science laboratories.
Mai previously co-founded humanoid startup Kind Humanoid, which was acquired by 1X in 2025. Bakhtiari previously worked as a Tesla Model 3 engineer and at one point reported directly to Elon Musk.
The company raised a previously announced $7.6 million pre-seed round led by Gradient Ventures. Gradient general partner Darian Shirazi said Feather has spent only a fraction of that financing so far.
Feather Prices Its Robot at About $30,000
Feather’s standard system currently starts at $29,990, considerably below some competing humanoid platforms. Mai told TechCrunch that the company is preparing for a larger product launch after addressing most of the problems found during field testing.
The startup’s founders want Feather to serve as an open platform that robotics developers can build businesses around, similar to how Android provides a common foundation for different mobile applications and devices.
Mai said Feather has taken inspiration from Chinese robotics manufacturers including Unitree, which has helped establish an early market for configurable humanoid hardware. He argues that a U.S.-built alternative could become increasingly relevant as restrictions affect the availability of newly imported foreign robot models.
Feather is betting that the number of companies developing specialized physical AI applications will increase substantially over the next five years. Rather than building every application itself, the startup plans to supply the hardware and development platform those companies can use for tasks across manufacturing, food service, laboratories and other workplaces.
Featured image credits: Linkedin
For more stories like it, click the +Follow button at the top of this page to follow us.