KOS AI, a health tech startup based in Silicon Valley, has created a non-intrusive glucose tracking tool that might revolutionize diabetes care globally for countless individuals. Following his key position as Lead for Machine Learning and Artificial Intelligence during the creation of the foundational technology used in the KOS Argus device, Stanford Computer Science alumnus Abhinav Agarwal has been named the new CTO of the company. In this capacity, he will oversee the ongoing growth of their AI-driven healthcare surveillance system.
The KOS Argus device employs optical sensing technology along with machine learning algorithms created by Agarwal’s group to gauge blood glucose levels via the skin without necessitating needle sticks or intrusive sensors. The firm claims that this tech boasts more than 91% precision in glucose tracking and shows 93.5% effectiveness in identifying hazardous hypoglycemic episodes.
This innovation marks a major step forward in preemptive health care via artificial intelligence," stated Abhinav Agarwal, who has recently been named as the Chief Technology Officer at KOS AI. "Our team created algorithms enabling the system to adjust to individual users' distinct biological characteristics. This could enhance precision progressively and uphold data confidentiality by handling all computations directly on the device.
This innovation arrives during a crucial period as the international diabetes surveillance sector keeps growing, projecting around 537 million adults globally affected by diabetes. Conventional blood sugar tracking techniques frequently cause discomfort which may result in decreased test rates and possibly deteriorated health conditions.
Under Agarwal’s guidance, the machine learning department at KOS AI developed an adaptable AI system that serves as the backbone of the Argus device. His group’s research into effective AI algorithms allows the gadget to execute intricate computations with minimal power consumption, which is essential for portable healthcare tech.
Casey Nguyen, who leads product strategy at KOS AI, emphasized the wider impacts: "The optical sensing and AI technologies we've created extend beyond just glucose tracking. They can be applied to various health indicators, possibly revolutionizing our approach to preemptive care."
Starting from the left, we have Casey Nguyen, followed by Abhinav Agarwal, and then Schalal Habibi.
During a press conference held on March 20, 2025, at their headquarters in Palo Alto, KOS AI provided further information regarding the technology. The firm showcased a live demo contrasting the KOS Argus with prevailing market offerings, emphasizing its non-intrusive method as opposed to conventional pinprick techniques.
At the demonstration, KOS AI exposed major discrepancies in conventional glucose monitors, indicating that tests taken simultaneously from the identical fingertip yielded a 30-40% disparity in results—an inconsistency beyond normal physiological limits. This contrast highlighted the possible benefits of the AI-powered optical sensing method created under Agarwal’s guidance for delivering steadier readings without needing blood draws or pricks.
The firm is presently carrying out clinical validation studies with an expectation of receiving FDA approval by late 2025. Priced at $349, this device aims to be a budget-friendly option compared to conventional continuous glucose monitors, which often have yearly expenses reaching into the thousands due to sensor replacements.
KOS AI was established by Schalal Habib, who owns five patents with an additional 45 under review. After earning his degree from Stanford’s Computer Science program, Agarwal brought his knowledge to the firm. Notably, his collaborative work with Casey Nguyen titled "Structured Pruning for Large Language Models" received acclaim at ICAIRME 2025. This expertise in enhancing machine learning processes significantly aided in developing the Argus device, enabling advanced algorithms to function efficiently within a compact wearable design.
About KOS AI
Based in Stanford Research Park in Palo Alto, KOS AI is a health tech firm that specializes in crafting AI-driven, non-intrusive health tracking tools. By integrating cutting-edge optics with sophisticated machine learning algorithms, they develop wearable gadgets for effortless health surveillance. Their mission involves producing systems that offer extensive wellness analytics without compromising individual privacy or convenience.
The post KOS AI Creates Non-Invasive Glucose Tracking Tech; Stanford AI Specialist Named Chief Technical Officer appeared first on Mom and More .