San Francisco Bay Area
As CEO of Niroga AI, I’m passionate about transforming healthcare through artificial intelligence, quantum sensing, and human-centered design. At Niroga AI, my focus is on AI-powered audio cardiography—creating accessible solutions that help monitor PTSD recovery, emotional well-being, and elderly family care through everyday devices. Before launching these ventures, I served over four years as Vice President of Research and AI at Sagence AI (formerly Analog Inference), where I led R&D in low-power AI hardware accelerators, computer vision, and generative AI. My work on analog-efficient inference, noise-tolerant algorithms, and quantization-aware architectures continues to shape the edge AI landscape. Across all my projects, my goal remains the same: to bridge cutting-edge science with compassion—building intelligent systems that extend human capability, deepen empathy in healthcare, and promote longer, healthier lives.
• As Founder and CEO of Niroga AI Inc., I lead the development of innovative AI solutions for Audio Cardiography. • Focused on addressing PTSD management and enhancing remote family care for the elderly. • Oversee the creation of biomedical devices and mobile applications tailored for cardiology-related needs. • Drive the vision of a startup dedicated to improving healthcare through advanced technology.
Responsible for research and development of essential technology, algorithms, system architecture and software for a Machine Learning and AI startup that is developing a low power Hardware Accelerator for ML Inference, using analog electronics and in memory compute. I have a strong background in software and algorithm design, and recently I developed and designed inference systems that minimize the computation power for Computer Vision and recent Generative AI LLM models. Invented high accuracy techniques for low bit width inferencing, and noisy calculations to preserve compute accuracy. I have a keen interest in advancing the field of generative LLM research.
Engineering and Technology development for Mobile and Connectivity eco-system • Senior Director, MediaTek; March 2015 – August 2020 (5.5 years) in the fields of Wi-Fi, Indoor Location, mmWave Radar, AI algorithms and design of energy efficient accelerated processors. I have developed several AI and classical algorithms for mmWave radar – Region of Interest detection, Human presence, Machine learning for pet/human discrimination, Pose and gesture detection and positioning in the indoor environment – from RF IQ data (Many mathematical techniques; ARM processors, energy constrained small processors, as host, e.g. Raspberry Pi etc. and use of CNN, RNN for solution to this problem). • A key insight I pioneered is application of the AI techniques for cancer detection from CT scans – and translating the solution to mmWave radar and human detection as a texture detection exercise. On its own merit I have personal algorithms that can detect lung cancer etc. with over 95% accuracy and come to top 10 ranks in the world in public dataset accuracy. • At MediaTek I started off with indoor location and Neighbor Aware Networking. To this problem as well we can apply AI analysis to CSI, V-matrix etc. I have had the opportunity for early participation and forming of 802.11az the next generation positioning studies. I pioneered Angle of Arrival in MIMO and TOF based estimation of location from 2012. Created several key patents for Wi-Fi neighborhood identification, and synchronization of multiple BSS in frequency and Time, for coexisting with Broadband (e.g. 40/80 MHz) and very narrowband (e.g. 1-2 MHz, but has much higher range). • At MediaTek, I also pioneered the work on Automotive Dead Reckoning (ADR) and used GPS and inertial sensors for improved location estimation in navigation applications and made MediaTek GPS chipsets more competitive and accurate. (Kalman filters, Integrating kinematics equations. etc.)
Broadband Video and Wireless Technologies • At Broadcom, since 2013-2015 I was a technical director, where I lead the efforts to develop ‘Indoor Location’ with Wi-Fi and engaged with potential customers such as Apple. The distance computations (with TOF) between 2-parties, as well as n-parties allow accurate and improved relative location aspects. Also pioneered how to use the Wi-Fi CSI, beamforming, and V-matrix for determination of distance. • In 2011-2013, I started the effort at Broadcom for Wireless charging, worked on out of band signaling for positioning, and helped develop A4WP standard etc. • I joined Broadcom Wireless BU in 2008 as an associate technical director, and from 2008-2012 I lead a software development team, and helped customize and modify Broadcom Wi-Fi for use in DTV. This was pioneering work. Some key aspects were lowering of latency below 10ms for video game applications (Nintendo consoles) etc. • From 2001-2008, I worked at the Digital Satellite and Broadband BU of Broadcom and was instrumental for driver, firmware and system development of several Digital Satellite chips and deployments to large customers such as Dish. Etc. I developed early elements of DVR and worked through the transition from Mpeg-2 to Mpeg 4 (AVC) and H.264. I lead the software and firmware team and was a senior principal scientist.
Technology for efficient object access and routing on the Internet • Founder and Chief Technology Officer, Object Exchange Inc. - October 2000 – September 2001 (1 year) - Developed pioneering technology for efficient object access and routing on the Internet, related to early cloud bandwidth exchange