BodhAssess — Phase 1

One Engine. Three Verticals. Every Psychometric Instrument.

Clinical. Industrial. Counselling. The same adaptive engine powers validated assessments across every psychology vertical — with IRT scoring, AI proctoring, and India-normed reports.

0 Calibrated Items
0 Indian Languages
0 Question Types
0 Access Tiers (T1–T5)

How It Works

From Item Bank to Signed Report

The BodhAssess pipeline — session creation through adaptive delivery, IRT scoring, and vertical-specific reports.

Create Session

Select instrument, respondent, language, and consent record. Batch CSV upload for cohort assessments. DPDP consent verified before any data collection.

Adaptive Delivery

CAT engine selects items via Maximum Fisher Information. Fixed-sequence for validated instruments (PHQ-9, GAD-7). Auto-save every 30 seconds. 11 languages.

IRT Scoring

MLE/WLE theta estimation after each response. T-scores (mean=50, SD=10) and percentile ranks against Indian population norms. Risk flags trigger instant alerts.

Report Generation

Vertical-specific PDF reports — clinical (DSM-aligned), industrial (competency profile), counselling (parent-friendly). Signed, stored in MinIO, DPDP-compliant.

The Engine

Built on IRT. Powered by AI.

100,000+ calibrated items. 3PL adaptive testing. India-normed scoring. Every capability the SRS demands.

3PL IRT Adaptive Engine

Computerised Adaptive Testing using the 3-parameter logistic model. Maximum Fisher Information item selection. SEM-based termination. Every score is a theta estimate with confidence interval.

100K+ Calibrated Item Bank

Every item carries IRT parameters (a, b, c), validated language versions, norm group data, and clinical risk flags. Items are never deleted — only deprecated. Full psychometric provenance.

AI Proctoring (YOLOv8 + MediaPipe)

Client-side YOLOv8-nano face detection, MediaPipe gaze tracking, Trust Score computation, and Secure Browser lockdown. BPaaS API available for third-party integrations.

4-Stage Report Pipeline

Score Computation (IRT + CTT) → Norm Referencing (Indian population T-scores, percentiles) → Template Rendering (vertical-specific jinja2) → Signed PDF generation. Clinical, Industrial, and Counselling report formats.

11 Indian Languages

English, Hindi, Tamil, Telugu, Marathi, Kannada, Bengali, Gujarati, Malayalam, Odia, Punjabi. TRAPD translation standard. Every instrument validated per language.

BodhLens Analytics

Natural language querying across all assessment and survey data. Ask questions in plain English — get AI-generated insights, charts, and cross-product analytics powered by Claude API.

Clinical Risk Flagging

PHQ-9 Item 9 (suicidality), self-harm indicators, and harm-to-others flags trigger instant alerts to the practitioner dashboard. Configurable threshold-based deterioration alerts on longitudinal tracking.

5-Tier Access Model

T1: Run library instruments. T2: Upload your own tool. T3: Full engine access. T4: Standardisation partnership with Bodh psychometric team. T5: Full AI integration with Claude API narrative generation.

BodhSurvey Integration

Survey creation with WhatsApp, SMS, and email delivery via Gupshup and Twilio. 11-language support. Response analytics integrated into BodhLens for cross-product insights.

Five Verticals. One Platform.

Four Doors. One Engine.

On login, you choose your vertical. The entire experience adapts — instruments, terminology, reports, AI features, and DPDP classification.

Clinical Psychology

For clinical psychologists, psychiatrists, and mental health practitioners. PHQ-9, GAD-7, DASS-21, Beck scales, PCL-5, AUDIT — all India-normed. MSE upload with AI-driven battery selection. ICD-10, ICD-11, and DSM-5 diagnostic coding in reports.

Industrial Psychology

For HR teams and I-O psychologists. Big Five (IPIP-NEO-120), HEXACO, Learning Agility, Cognitive Aptitude Battery, SJTs, and the AI Adaptability Index — 7 dimensions measuring readiness for AI-augmented work. AI proctoring by default.

Counselling & Child Psychology

For school counsellors and child psychologists. Age-normed SCAS, CDI-2, ADHD Rating Scale-5, Developmental Milestones Tracker, Academic Stress Inventory. Multi-informant design with parent and teacher parallel forms. Parental consent workflow built in.

Designing Experiments

For researchers and clinical scientists. IAT, Dot Probe, Stroop, Go/No-Go, N-Back, Delay Discounting — delivered via jsPsych with millisecond timing precision. Trial-level raw data export. Combine implicit paradigms with self-report in a single session.

White-Label (Meta-Vertical)

Schools, hospitals, and consultancies run BodhAssess under their own brand. Custom domain, logo, report letterhead, and terminology. End users never see the Bodh name. The underlying engine — CAT, proctoring, BodhLens, AI scoring — is identical.

BodhProctor — AI Proctoring

Assessment Integrity Without a Human Proctor

YOLOv8-nano object detection, MediaPipe gaze tracking, and Trust Score computation — running client-side for zero latency. Available as BPaaS API for third-party platforms.

YOLOv8 + MediaPipe Detection

Real-time face detection, multi-person alerts, gaze direction tracking, and phone/object detection. All processing happens client-side — zero server dependency.

Secure Browser + Trust Score

Full-screen lockdown, copy-paste disabled, tab-switch logging, screen capture blocked. Trust Score computed from all anomaly events — delivered in the final Trust Report.

BPaaS — Proctoring as a Service

Third-party platforms integrate BodhProctor via REST API. Create proctoring sessions, receive WebSocket events, retrieve Trust Reports. Your platform, our integrity layer.

Secure assessment proctoring

Compliance & Standards

Psychometric Rigour. Regulatory Compliance. By Design.

Every instrument, every data point, every report meets the highest psychometric and regulatory standards.

DPDP Act 2023

Full compliance — consent management, erasure pipeline, data residency on Indian soil (BLR1/BOM1), Data Principal Rights Portal, and DPO appointed.

ICD-10 / ICD-11 / DSM-5

Clinical reports present diagnostic codes from India's ICD-10 standard, WHO's ICD-11, and APA's DSM-5. Severity classifications and risk indicators mapped to diagnostic categories.

IRT 3PL Validated

Every adaptive instrument uses the 3-parameter logistic model with published item parameters. Indian population norms per instrument. Confidence intervals on every score.

Multi-Tenant RLS

PostgreSQL Row-Level Security isolates every tenant. Keycloak JWT auth with 15-min expiry. MFA mandatory for admins. SSO via SAML 2.0 and OIDC. Account lockout after 5 failed attempts.

The Assessment Engine
India Has Been Waiting For

Clinical. Industrial. Counselling. Experimental. White-Label. One IRT-powered platform for every psychology vertical.