Risk & Quant Analytics
Developed and validated models for risk scoring and forecasting; partnered with stakeholders to translate model outputs into decisions. Reduced false positives and improved stability across market regimes.
Data Science & Applied Machine Learning
I design and deploy data-driven solutionsâfrom timeâseries modeling and NLP to risk analyticsâthat turn complex datasets into business decisions. This page intentionally avoids personal contact details; use the button above to request them securely.
Applied ML engineer and quantitative problemâsolver with experience across highly regulated industries. I partner with crossâfunctional teams to frame problems, ship models, and measure impactâbalancing statistical rigor with production constraints.
This site highlights representative work and skills while deliberately omitting phone, email, home address, and other personal identifiers. If you need those, use the secure request flow.
Developed and validated models for risk scoring and forecasting; partnered with stakeholders to translate model outputs into decisions. Reduced false positives and improved stability across market regimes.
Shipped NLP and timeâseries services behind APIs with monitoring and A/B testing, cutting model latency and improving downstream KPIs.
Delivered data products across finance, insurance, telecom, and techâbalancing compliance, model governance, and user needs.
(Listed for context only; this section intentionally omits dates, locations, and contact details.)
Endâtoâend timeâseries pipeline (feature store â model registry â monitoring) that improved forecast MAPE and reduced incident load.
Built classification & extraction on unstructured text with embeddings and weak supervision; exposed via REST API.
Templates & checks for documentation, fairness, and drift to streamline approvals in regulated environments.
Why cost curves, calibration, and decision metrics beat leaderboard chasingâespecially in production systems.
Lightweight patterns to move from notebooks to stable services without overâengineering.
This site intentionally excludes email, phone number, physical address, and personal identifiers. Any request for contact details is handled via a consentâbased flow. Analytics are disabled; no cookies are set.