From the infrastructure layer.

AI Fairness in Credit Scoring: DTPG, RVDR, and CFDI Explained

Fairness in lending AI is a behaviour, not a property. Learn how three operational metrics — DTPG, RVDR, and CFDI — make it measurable across LightGBM + HDBSCAN architecture, and why conventional MLOps monitoring misses all three.

Recourse Is Not an Apology — It Is Infrastructure

Actionable recourse is a technical system requirement, not a communication gesture. Its design, validity, and lifecycle must be engineered and monitored with the same rigour as the model itself.

Why Your Credit Bureau Score Is a Rearview Mirror

The bureau score is high-precision for the population it was built on. For 451 million Indians with thin or absent tradelines, it structurally cannot evaluate risk well. Here is the technical case for governed multi-signal decisioning — and why governance is the piece the industry keeps skipping.

What the RBI FREE-AI Framework Actually Asks of You — And What Most Lenders Are Getting Wrong

FREE-AI is not a policy checklist. It is an operational governance requirement demanding demonstrable, auditable, continuously monitored controls. Here is what each of the six responsible-AI dimensions actually requires, where institutions are failing, and what technically adequate looks like.

The Audit You Are Not Ready For

Regulatory audit readiness in AI-driven lending is not a documentation exercise. It is an evidentiary infrastructure problem — and most institutions are structurally unable to answer the questions regulators are beginning to ask.

What ₹65+ Lakh Crore in NBFC Lending Has to Do With Your Model's Threshold

Threshold placement is the highest-leverage economic decision variable in a lending book — yet most NBFCs treat it as a tuning parameter, not a governed policy decision. A quantitative walkthrough of threshold economics, governance, and segment-stratified calibration.

Thin-File Borrowers and the Next Frontier of Lending Growth

India's largest untapped credit opportunity is not a new product. It is the governed deployment of existing alternative data infrastructure to serve a population that is economically active but credit-invisible under traditional scoring.

The Explainability Problem Nobody in Indian Lending Is Talking About

Explainability in lending AI has crossed from a model-development concern into a governance-infrastructure requirement. Most Indian lenders are only solving the easier half — generating explanations — while ignoring the harder problem of preserving, auditing, and monitoring them across a model's lifecycle.