Governance that works · Evidence-first · No advertising tracking
Technology risk. Evidence. Resilience.
A practitioner's journal at the intersection of audit defensibility, control automation, and AI governance, for regulated environments that demand evidence, not promises.
// Personal writing and small builds, published in my own time. Views are my own.
// navigate
The Seven Dhārās
Seven streams of practice. Each one a discipline.
Dhārā I
स्वामी
GRC Intelligence
Policy-as-code, control automation, evidence engineering.
Dhārā II
अमात्य
AI-Augmented Governance
Minimum sufficient control, maturity radar, direction scoring.
Dhārā III
जनपद
Economic Statecraft
Geopolitical economics as a governance signal engine.
Dhārā IV
दुर्ग
Resilience Engineering
Systems that hold under adversarial pressure, recover by architecture.
Dhārā V
कोश
Signal Sovereignty
Weak signal convergence. Niti-driven capital discipline.
Dhārā VI
दण्ड
Regulatory Cartography
Cross-jurisdictional signal engine: what shifts, where, when.
Dhārā VII
मित्र
Sovereign Judgment
Permission architecture. Vendor maturity. Programme design.
// field notes
Latest Field Notes
Recent dispatches. Each one lives inside its Dhārā.
AI-Augmented Governance
Open weights solve vendor opacity. They do not solve training-data provenance. Corpus integrity, RAG provenance, and prompt injection are the gaps your AI governance programme is missing.
AI-Augmented Governance
Trust primitives are technical artefacts. Whose authority chain backstops the hash, the key, and the pin? A supply-chain decision framework for self-hosted AI in regulated firms.
AI-Augmented Governance
Why self-hosted open-weights inference is the boardroom-defensible default for regulated firms, and where hybrid still makes sense.
AI-Augmented Governance
Regulated records can outlive the cryptography that proves them. Crypto-agility is the property that lets a firm migrate primitives without re-attesting the world. The series closer.
Resilience Engineering
MCP runs every connected server inside your model's trust boundary. The host, not the model, is where integrity holds, and the failure modes are fixed by architecture, not vigilance.
AI-Augmented Governance
Hardware attestation for self-hosted AI in regulated firms. The firmware layer is unscoped, the BMC is unpatched, and the DORA register excludes the GPU vendor.