How it works — end to end.
On-device computer vision. 7 clinical frameworks. Certified ergonomist validation. One personalised report.
Assessment in progress
DPDPA compliant · Face-blurred on-device
All the tools to evaluate workplace ergonomics
A worker assessment on any phone — camera or camera-free — and an aggregate risk dashboard for the employer, from one clinical engine.
7 frameworks live, 8 coming
Each assessment opens with the NMQ discomfort survey, then scores RULA, ROSA, REBA, NIOSH, OCRA, JSI and OWAS — with LUBA, EAWS, MMC-ISO and more on the roadmap.
On-device privacy
Faces are blurred on the device and full video is never uploaded.
Reports, two ways
A personalised worker PDF and an aggregate employer risk dashboard.
One multi-method engine
Add a method and it inherits the whole pipeline — capture, scoring and report.
7 frameworks live, 8 coming
Each assessment opens with the NMQ discomfort survey, then scores RULA, ROSA, REBA, NIOSH, OCRA, JSI and OWAS — with LUBA, EAWS, MMC-ISO and more on the roadmap.
On-device privacy
Faces are blurred on the device and full video is never uploaded.
Reports, two ways
A personalised worker PDF and an aggregate employer risk dashboard.
One multi-method engine
Add a method and it inherits the whole pipeline — capture, scoring and report.
From phone to personalised report in 5 steps.
No hardware. No app install. No manual scheduling. A complete clinical assessment that fits in a browser tab.
Worker receives link or QR
No app install. Opens in any mobile browser. Works on any smartphone.
Captures front + side views
~2 minutes. Camera switches off after capture. Video never leaves the device.
AI scores posture on-device
33 body landmarks extracted. Joint angles computed geometrically. 7 frameworks scored — all on the device, before anything is sent.
Ergonomist reviews and validates
AI output goes to a certified ergonomist. They review scores, edit sub-scores, add clinical notes. Nothing ships un-reviewed.
Worker accesses OTP-gated report
8-digit code emailed. Personalised PDF with prioritised recommendations.
Step 2 — Worker captures workstation
Joint angles computed by geometry from on-device landmarks · ergonomist-validated · face-blurred on-device.
Step 3 — AI scores on-device. Real output.
What the AI actually measures.
Pure geometry from landmark coordinates — not self-report, not manual goniometer.
All angles computed client-side using dot-product + arccos on 2D pose-landmark coordinates. No server round-trip for scoring.
Joint angles computed by geometry from on-device landmarks · ergonomist-validated · face-blurred on-device.
Seven frameworks. One engine.
Each method built from its primary-literature blueprint. No copy-paste.
Assess exposure to risk from whole-body postures, including lower extremities — particularly for tasks involving unpredictable, varied postures.
Assess exposure to risk due to posture, duration, frequency, and force for the upper limbs, neck, and trunk — particularly in sedentary, computer-intensive, or precision work.
Evaluate ergonomic risk at computer workstations — assessing chair, monitor, keyboard, mouse, and telephone configuration against validated office ergonomics standards.
Calculate the Recommended Weight Limit (RWL) for two-handed manual lifting tasks and determine the Lifting Index (LI) — the ratio of actual load to the RWL.
Quantify upper-limb exposure to repetitive, high-frequency manual work and predict the risk of work-related musculoskeletal disorders of the shoulder, elbow, wrist, and hand.
Assess the risk of distal upper extremity musculoskeletal disorders — specifically the wrist, hand, and forearm — in jobs involving repetitive or forceful exertions.
Analyse the distribution of working postures across an entire work cycle through time-sampling — identifying what proportion of time is spent in each posture category.
A posture-only joint-motion assessment focused on physical discomfort and localised stress accumulating in the trunk, neck, and upper limbs.
A comprehensive whole-body screening worksheet that evaluates working postures, action forces, manual material handling, and repetitive upper-limb load in a single instrument.
An international standardised compliance protocol establishing ergonomic limits and risk estimation for manual lifting, lowering, and carrying.
A psychophysical design reference for the safety of manual handling tasks — including pushing, pulling, and carrying alongside lifting and lowering.
A physics-based modelling approach that calculates internal mechanical stresses — compression and shear forces — acting on muscles, joints, and spinal discs during work.
A comprehensive, multi-variable ergonomic audit that evaluates physical strain, environmental factors, and mental workload simultaneously.
A practical, action-oriented workplace screening checklist designed to identify high-impact improvements across a wide range of operational conditions.
A thermal-comfort model that evaluates and predicts how large groups of people experience the climate of an indoor space.
Methods marked Soon are on our roadmap — each new method costs ~20–30% of the first on the shared engine.
Four ways to deploy — for every workforce context.
From a 5-minute phone capture to a structured text questionnaire. Configure per campaign.
Worker films their workstation on any phone.
On-device AI scores posture from video. Fastest time-to-report.
Structured digital questionnaire with illustrated questions.
No camera required. Scored against the same clinical frameworks, every result ergonomist-validated — built for sites where cameras aren't allowed.
Combines video capture with guided questionnaire.
Richer data for complex tasks. Best signal-to-noise for ergonomist review.
Runs multiple methods in one campaign.
A single worker completes one assessment that generates RULA + ROSA + NIOSH scores simultaneously.
AI computes. The ergonomist signs off. Every time.
What they see
- AI scores per method
- Joint angles & confidence
- Pose photos (face-blurred)
- Pain diagram (NMQ)
What they edit
- Sub-scores per body segment
- Recommendations & priority
- Clinical notes
- Risk narrative
Approval gate
Approve is blocked until all methods are validated. Every score is signed off before the worker sees it.
“The clinical methodology, framework selection, and recommendation logic the platform encodes are Dr. Mona Pankaj’s work — not a software team’s interpretation.”
Privacy by design — three mandatory layers.
Face blur is permanent. Body blur is configurable. Render opacity is configurable. None can be removed by the employer.
Layer 1 — Face blur
An elliptical pixel block is drawn over the detected nose landmark on every captured frame. Cannot be disabled. Applied before any data leaves the device.
Layer 2 — Body blur
0–100% whole-canvas pixelation per campaign. Configured by your Super Admin. Campaigns can only tighten — never loosen — the company default.
Layer 3 — Render opacity
How faintly the photo shows behind the skeleton wireframe in the report. 0% = skeleton-only. 38% = default overlay. SA-controlled per campaign.
Compliance
Captured on any phone · processed on-device · the source video never leaves the worker's device.
Built for enterprise deployment from day one.
Security, compliance, and clinical rigour aren't features — they're the foundation.
Multi-tenant
Employer sees aggregates only. Clinical data is SA + Ergonomist exclusively — enforced at the database row-level.
No app install
A link or QR code. Any smartphone browser. No MDM, no app store, no IT provisioning.
SSO ready
OAuth2/OIDC on the roadmap → SAML for enterprise SSO. No blocker for enterprise deployment.
BRSR-mappable
Every assessment generates OHS data that maps directly to SEBI BRSR reporting requirements for listed companies.
Full audit trail
Every ergonomist action is logged with timestamp, identity, and change record. Legally defensible, ISO 45001 aligned.
IT requirements
No firewall changes for employees. Data stays in Mumbai region. Read the full IT spec before onboarding.
Ready to see it with your team?
Request a pilot — 20–50 assessments, clinically validated reports, employee feedback before you roll out.