Sense
Capture raw wearable signals alongside task events and participant reports, using a consented study protocol.
From raw signals
to personal understanding.
ManasAI is building a research-driven platform that connects wearable sensing with customised wellness models. Labs come first, followed by wellness centres working with defined research protocols.
Discuss your workflow ↗The architecture
Sense → Infer → Guide → Learn
A development framework for turning observations into carefully evaluated, useful feedback.
Infer
Develop features and models, personalise against a baseline, and evaluate uncertainty and generalisation.
Guide
Explore feedback with researchers and participants. Ground it in evidence and preserve individual choice.
Learn
Use study findings to refine protocols, models, and hardware under agreed data-governance terms.
Priority applicationsTwo starting points.
Two starting points.
Individual baselines.
Wellness AI should account for differences between people, environments, and sessions. Our immediate focus is on the questions below.
Stress & anxiety
Investigate arousal and recovery using physiological signals, contextual events, and self-reports. Develop models that distinguish an individual’s patterns rather than assigning a universal stress score.
Attention & distraction
Study focus, restlessness, and task disengagement using movement and physiological data alongside attention tasks and participant feedback. Evaluate what the signals can reliably tell us.
Research applications, not diagnostic services. Performance must be established for the intended protocol and population.
Technical directionA foundation you can
A foundation you can
build on.
Physiological AI
Signal quality, artefact handling, and multimodal modelling, with physics-informed and physiology-aware approaches where appropriate.
Behavioural analytics
Task annotations, reaction times, and self-reports to give physiological observations a meaningful study context.
Custom models
Research workflows for individual baselines and adaptation, with evaluation across participants and sessions.
Research datasets
Study-specific plans for consent, access, retention, and sharing. Data use and model-development rights should be agreed before collection.
Longer-term researchRoom for the wider
Room for the wider
science of wellbeing.
These directions remain part of the vision. They are not separate products available for purchase.
Sleep & recovery
Investigating rhythm, routine, and recovery over time.
Meditation
Relating stillness, breath, and reported experience without grading the quality of practice.
Reflective guidance
Exploring culturally grounded retrieval and dialogue over Indic knowledge, with sources and evaluation.
Your protocol, our starting pointBuild the study
Talk to the team ↗Build the study
before the score.
Let’s define the data and evaluation your research needs.