Services

Practice areas

The areas change; the criterion does not. In all of them delivery includes the model, the code, documented assumptions and the validation evidence behind them. A good part of the work ends up being backend: the model has to land in the hands of the desk or team that will use it, not in a document.

Statistics and machine learning

Predictive models built to be valid, not merely to fit well in sample. This is the area where apparent performance is most often mistaken for real performance.

Predictive modeling and model selection on explicit criteria, not by habit.
Uncertainty quantification: intervals you can actually decide on.
Validation design: out-of-sample, by horizon, with information leakage controlled.
Independent review of third-party models, machine learning ones included.
Deployment with drift monitoring and reproducible retraining.

Mathematical finance and derivatives

Pricing, calibration and numerical methods for volatility surfaces and nonlinear structures. This is day-to-day work on the team.

Optimal-transport-based local volatility calibration: arbitrage-free, stable and real-time.
Surface calibration: parameterization, smoothing and no-arbitrage constraints.
Numerical schemes: finite differences, quadrature, Monte Carlo and transform methods.
Independent review of existing models and their discretization error.
Pricing libraries with regression tests and analytic benchmarks.

Risk, heavy tails and statistical arbitrage

Risk measurement where Gaussian assumptions fail — which is precisely where risk matters.

Tail index estimation and extreme value fitting on real data.
CVaR limits posed as a convex problem, optimizable jointly with position size.
Design and evaluation of statistical arbitrage strategies, costs and capacity included.
Honest backtesting: out-of-sample, with transaction costs, without overfitting the period.
Portfolio risk: congestion, basis, hourly shape and scarcity events.

Power and energy markets

Nodal price, demand and risk models for participants in the Mexican power market and international counterparties.

Locational marginal prices: modeling, decomposition into energy, losses and congestion, and forecasting.
Load and consumption forecasting across horizons.
Valuation of hedges, PPAs and physical positions.
Import and export strategies, and their exposure to scarcity risk.

Data, backend and productization

Most of the work ends up here: the model has to run on its own and sit within reach of whoever decides.

Automated acquisition: scrapers, public APIs and market feeds.
Transformation and validation pipelines with quality controls and alerting.
Schema and database design for time series and market data.
APIs, scheduled jobs and monitoring so the desk queries the model without an intermediary.
Dashboards and internal applications on one source of truth.

Actuarial modeling

Valuation of contingent liabilities, decrement tables and stochastic projections with explicit assumptions.

Valuation of annuities and long-dated obligations.
Construction and fitting of mortality, morbidity and persistency tables.
Stochastic projection and reserve adequacy testing.
Assumption documentation fit for regulatory or audit review.

Expert analysis and litigation support

Quantitative analysis for legal proceedings, written to be understood by non-mathematicians.

Probability of non-payment and valuation of disputed cash flows.
Damages quantification and counterfactual scenarios.
Technical critique of opposing models.
Reports and exhibits with traceable assumptions, sources and sensitivities.

Ways to work together

Scope is defined before work starts. So are price and deliverables.

Bounded diagnostic

One to two weeks to state the problem properly, review what exists and propose a route. Ends in a document with options, effort and risks.

Fixed-scope project

Defined scope, acceptance criteria and schedule. Delivers the model, code, documentation and a handover session.

Ongoing quantitative support

Recurring availability for modeling, independent review and development, at whatever load the work requires.

Expert opinion and testimony

Analysis and testimony for legal or arbitral proceedings, with reports prepared to withstand cross-examination.

Have a problem that spreadsheets won't settle?

Get in touch