model-risk-management

Synapsa Commons: free, hands-on AI courses that run anywhere (Colab, Kaggle, Binder, Codespaces, Jupyter). Build EU AI Act conformity evidence, validate models like a risk committee, price predictive-maintenance alarms, measure document extraction, and make a simulated humanoid walk. Every lesson autograded. From the team building Synapsa.

FLUXARA-1 independent audit verification pack for peer review. Contains frozen engine outputs, replication scripts, and vBase blockchain temporal proofs (SR 11-7 / Basel IRB aligned).

Automated validation toolkit for tabular ML models in finance and regulated domains.

Experiments using LLMs for whitepater analysis from the AI ML risk management perspective

Governed multi-LLM Responsible AI control plane for prior-authorization decision support, with PHI-safe audit trails, HITL review, evidence packets, and deterministic governance evals.

Enterprise-grade credit risk platform on the real Home Credit Default Risk (Kaggle) dataset -- 5 hardened Mega Projects, 30 real verified notebooks, 15 deployable FastAPI services, real statistical validation and explainability throughout. Zero fabrication: every number is computed live, never asserted.

Institutional Stock Risk Engine & Governance Framework

Independent model-risk validation of a consumer credit default model to SR 11-7 standards: a gradient-boosted champion tested against an independently built benchmark.

Interactive ML model validation framework: explainability, weak spots, robustness, fairness — widget-driven, no-code demo

AI Governance portfolio case study assessing a high-risk automated loan underwriting system using VerifyWise, EU AI Act, NIST AI RMF, and ISO/IEC 42001 concepts.

Checks a derivative pricing model against closed-form, lattice, PDE and Monte Carlo reference engines, then reports what is wrong in plain language. Vanilla and exotic options; Black-Scholes, Heston and local volatility.

AI Governance Portfolio for an Automated Loan Underwriting System

AI Governance portfolio case study assessing a high-risk automated loan underwriting system using VerifyWise, EU AI Act, NIST AI RMF, and ISO/IEC 42001 concepts.

Institutional-grade LLM evaluation, red-teaming & telemetry framework for financial AI. Tracks 56 AI ecosystem tickers with an 18-kernel audit architecture; 10 production auditors live.

Experimental AI oversight components for banks and insurers: drift, explanation consistency, fairness metrics, policy evaluation and accountable review.

Credit decisioning with model risk governance on 1.96M SBA loans and 93.4M HMDA applications. Ten promotion gates — one blocks my own model.

A monitoring layer for a model I built, checking whether it's still trustworthy months later. Found a real blind spot in the drift detection, caught a false model promotion before it shipped, and a subgroup finding that looked unfair until I checked the right metric. Built around the PRA's actual model risk principles.

Independent model validation and AI governance framework for high-stakes financial decision models using public HMDA mortgage data.

raucle gives every AI agent action a cryptographic receipt: what it did, by whose authority, against which provably-enforced policy. Portable across frameworks and organisations — signed trust registry, agent passports, cross-org handshakes, compliance evidence packs.

Top Responsible AI Platform (Opensource) 🌟 Star if you like it! 🌟