AI Audit Lab — 12 frames across ISO 42001, 27001, 27701, DPDP 2023, SOC 2, NIST AI RMF
The Lead Auditor of artificial intelligence.
Dr. Sodhi · Triple-ISO Lead Auditor · ISO/IEC 42001 · 27001 · 27701 (ANSI/ABICB, Mar 2025) · Architect of the audit engine
1-hour readout: which findings block launch, which are theatre, exactly what to fix this week — plus a signable PDF for the board on ISO 42001 / 27001 / 27701.
Audit engine online. Pick a frame
// 12 audit frames · ISO 42001 / 27001 / 27701 signable by Dr Sodhi · from ₹799 · pay per run.
Engine: NeuroCortex v2 →
// Other pillars: → Human Lab · → Business Lab
12 standards live · ₹799 each. ISO 42001 / 27001 / 27701, SOC 2, DPDP, GDPR, NIST AI RMF, BIS/STQC, ISO 9001, HIPAA, CDSCO, HITRUST.See the full menu →
// 1. PICK AN AUDIT FRAME
12 frames live · flat ₹799 per standard. Tier-1 ISOs + DPDP + ISO 9001 can be upgraded to a Dr. Sodhi-signed edition after scoring.
// 2. INTAKE · ISO 42001
₹799 per frameAnswer each field as specifically as you can. The engine uses these to write 16 auditor-grade questions tailored to ISO 42001.
// PAST REPORTS
Essays from the AI Audit Lab.
AI audit in India — the complete 2026 guide (12 frames, 4 tiers)
What an AI audit actually covers, the 12 frames an Indian company can pick from (ISO 42001, NIST AI RMF, EU AI Act, DPDP, OWASP LLM Top 10 and 7 more), the 4 tiers (₹799–₹24,499), and the 3 controls Indian firms fail most.
AI governance framework — a practical 2026 template for Indian enterprises
Six-layer AI governance framework — policy, roles, inventory, risk, controls, review — mapped to ISO 42001 and NIST AI RMF, with the 90-day build that actually ships.
LLM evaluation framework for production — beyond accuracy in 2026
Seven-axis LLM evaluation — correctness, faithfulness, safety, latency, cost, robustness, fairness. Why public benchmarks mislead, how to build a real eval set, and when LLM-as-judge can be trusted.
AI red teaming — what to probe, what to document, what to fix
Four-layer AI red team — capability, safety, security, abuse. Reproducible findings register, India-specific attack vectors (code-mix jailbreaks, caste/religion abuse), and the two-week engagement structure.
The Lead Auditor of artificial intelligence.
Dr. Sodhi · Triple-ISO Lead Auditor · ISO/IEC 42001 · 27001 · 27701 (ANSI/ABICB, Mar 2025) · Architect of the audit engine