Deeptech startup NeuroDx has unveiled MANAS‑1, a 400‑million‑parameter brain foundation model that uses AI to decode human brain electrical activity and support early diagnosis of neurological disorders. The model was launched at the India AI Impact Summit and is among the first India‑built AI systems focused specifically on clinical‑grade brain health diagnostics. NeuroDx is one of 12 startups selected under the IndiaAI Mission, signalling strong institutional backing for its approach.
Native AI Model Built on Indian Clinical Data
MANAS‑1 has been developed using Indian clinical data and research, positioning it as a fully native model aligned to domestic healthcare realities rather than imported benchmarks. The system is designed as a decision‑support tool for neurologists, psychiatrists and researchers, offering scalable and ethical insights into brain health. NeuroDx is targeting the growing global burden of neurological and psychiatric conditions, which affect over 600 million people worldwide, with a particular focus on improving speed and accuracy in diagnosis.
At the core of the platform is transformer‑based analysis of electroencephalogram signals and other neurological data. According to the company, MANAS‑1 has achieved accuracy levels above 95 percent for identifying biomarkers in epilepsy and select mental health disorders. The model is intended to augment clinical judgment, helping specialists detect patterns that may be difficult to observe consistently through manual interpretation alone.
Deeptech Team Targets Affordable, Non‑Invasive Diagnostics
NeuroDx was founded by Kailash Sati and Siddharth Panwar, who bring experience from AIIMS New Delhi, Stanford University, IIT Delhi and IIT Kanpur. The company is focused on building affordable, non‑invasive diagnostic tools powered by AI, integrating human expertise with model‑driven insights to improve clinical confidence and research outcomes. By combining neuroscientific domain knowledge with AI engineering, the founders aim to reduce diagnostic delays that often worsen patient outcomes in neurological care.
In parallel to MANAS‑1, the startup is developing a significantly larger 20‑billion‑parameter foundation model dedicated to EEG signal analysis. This next‑generation system is expected to support early screening of neurological conditions at scale and enable future brain‑computer interface applications in areas such as assistive technologies and neurorehabilitation.
Part of India’s Emerging Sovereign Healthcare AI Stack
With MANAS‑1, NeuroDx joins a growing group of Indian startups working on sovereign AI systems for critical sectors like healthcare. The push aligns with government efforts under the IndiaAI Mission to build domestic AI capabilities in strategic domains, reduce dependence on foreign models and ensure that sensitive health data remains within national jurisdiction. As more clinical partners and research institutions adopt MANAS‑1, the model’s performance is expected to improve further, reinforcing India’s role in the global neuro‑AI landscape.
