On April 2, 2026, Epia Neuro, a brain-computer interface startup headquartered in San Francisco, officially concluded five years of stealth research and development and announced its establishment. The company focuses on developing a novel implantable neural interface platform, aiming to help stroke survivors and patients with cognitive decline restore hand function and life independence.
It is reported that the founding team of the company is led by Michel Maharbiz, a serial entrepreneur, serving as the CEO. Maharbiz was previously the founder of Iota Biosciences and is currently a professor at the University of California, Berkeley. The Chief Technology and Science Officer is Gil Mandelbaum, a neuroscientist with over 15 years of experience in scaling high-growth enterprises. Notably, the team has been operating in stealth R&D mode for the past five years, focusing on breakthroughs in the foundational technology of their product.
Epia Neuro's core product is a "read-write" minimally invasive neural interface. The disc-shaped implant is placed within the skull, and the surgery takes only one hour, without penetrating the dura mater, making future upgrades and replacements relatively convenient.
The platform supports both cortical and deep brain sensing and electrical stimulation functions, capable of reading neural intentions and compatible with cortical stimulation and deep brain stimulation therapies. During patient rehabilitation or at home, an AI-driven power-assisted glove needs to be worn. The system wirelessly charges the implant via a non-invasive headpiece, and its built-in algorithm combines neural signals with sensor data to drive grasping motions in real time, achieving an "intention-action" closed loop.
Epia Neuro Product Demonstration Diagram
This technology is based on the principle of neuroplasticity. After a stroke damages the motor cortex, the patient's brain can still generate motor signals, but the injury blocks the transmission of these signals to the muscles. Epia's implant collects signals from undamaged areas, identifies intent, and translates them into actual movements via a glove. With repeated use, the system strengthens motor-related neural pathways, inducing brain plasticity, ultimately allowing patients to genuinely improve their native functions even after removing the glove.
Unlike Neuralink, which primarily helps patients control computers, Epia focuses on assisting patients in regaining movement of their own hands. Compared to the previously FDA-approved wearable device IpsiHand (which some patients cannot use due to insufficient recognizable brain signals), Epia’s implantable solution is closer to the brain and can extract more meaningful signals. According to the company, the first human technical demonstration is planned to take place later this year at Lenox Hill Hospital in New York, with further testing expected before the end of 2026.
Notably, Epia's technical approach shows a high degree of consensus with Neuracle, a Chinese brain-computer interface company. Just one day before Epia's announcement of its establishment, the National Medical Products Administration officially released Neuracle's registration technical review report. The report confirms that Neuracle’s signal acquisition through epidural electrodes demonstrates that a minimally invasive approach can fully obtain high-quality signals sufficient to support precise motor decoding.
Neuracle Product Demonstration Diagram
In terms of external effectors, Neuracle has achieved a closed-loop system for brainwave intention-driven physical peripherals by wirelessly transmitting signals to drive a dedicated pneumatic glove. In terms of clinical data, the results of Neuracle's multi-center trials show that after six months of follow-up with subjects, the response rate for brain-computer interface-assisted grasping reached 100%, the average accuracy of decoding algorithms stabilized at 83%, and there were no cases of device-related serious adverse events.
Overall, Epia has officially entered the market after five years of covert research and development, bringing the imagination of a two-stage intervention model. The review data recently disclosed by Neuracle provides strong logical support for this technical route. Whether Epia, as a latecomer, can replicate or even surpass this data performance in its first human trial will be the key to determining whether its product can fulfill its ambition.
