Machine learning for financial risk management
A framework for estimating Basel IV capital requirements.
A systems-thinking essay arguing that most optimization quietly trades away buffers, slack, and resilience to make present metrics look better. It reframes efficiency as borrowing stability from the future, and shows how education, workforce, infrastructure, markets, and hardware all get optimized into fragility.
The repo contains the main topics carried out in my master's thesis on operational risk. In particular, it is described how to implement the so called Loss Distribution Approach (LDA), which is considered the state-of-the-art method to compute capital charge among large banks.
Risk-based SLA compliance and remediation pressure monitoring framework built in Power BI with custom prioritization logic.
Living freight-fraud investigation simulator: autonomous freight operations, behavioral signals, false positives and risk investigations — with simulation ground truth hidden from the analyst.
Explainable operational risk intelligence for multi-stage supply chains, combining geopolitical risk signals, D1–D4 risk propagation, counterfactual analysis, and human-in-the-loop decision support.
Developed an AI agent to monitor operational risks and performance metrics. Integrated multi-source data using ETL pipelines, applied ML for anomaly detection and risk prediction, and generated AI-driven insights through automated alerts and dashboards. Containerized the solution using Docker for scalable and efficient deployment.
RISK//OS — an open-source Program Risk & Delivery Operating System
Build AI-driven crypto trading infrastructure with compliance-first tools for trusted digital finance
NIST SP 800-34 aligned Business Continuity Plan and Risk Profile for FinTech organizations. BIA, RTO/RPO definitions, dependency mapping, critical process analysis, and resilience strategies for governance and audit readiness.
A quantitative framework for modeling Operational Risk Capital under Basel III standards using the Loss Distribution Approach (LDA). Implements Monte Carlo convolution of Poisson frequency and Generalized Pareto (Heavy-Tailed) severity distributions to calculate the 99.9% Value at Risk (VaR).
End-to-End data engineering pipeline (Python/PostgreSQL) and interactive Power BI dashboard designed to identify compliance risks and automate financial anomaly detection.
Risk Analytics audit of EU DSA transparency reports, translating 359.85M enforcement measures into a human-review and quality-governance framework.
Operational risk Monte Carlo (Poisson/Lognormal) for collision losses—methods, R code, and 99.9% capital estimate.
LDA probabilistic risk profiling — latent risk archetypes, portfolio mix drift, book-transfer segmentation
Operational Risk and Reliability Analytics: Python, SQL, and Power BI system for failure probability, severity, utilization thresholds, reliability analytics, and cost exposure.
A practical checklist for identifying fragile Excel spreadsheets, structural risks, and reliability issues in operational workflows.
Power BI complaint risk dashboard for executive reporting, timeliness monitoring, company benchmarking, drill-through, and governance.
Streamlit dashboard monitoring 300 vendor contracts - composite risk scoring across SLA, data issues, and expiry urgency, 30/60/90-day renewal flags, root cause tagging, and 3-sheet Excel action reports. Live app deployed on Streamlit Cloud.
To associate your repository with the operational-risk topic, visit your repo's landing page and select "manage topics."