Some problems are clean enough to belong to one field. The interesting ones usually are not.
I take on consulting work where mathematics, statistics, software, artificial intelligence, or public policy have to produce an answer somebody can actually use. That may mean building a model, reviewing one, making sense of ugly data, designing an analytical system, estimating risk, comparing decisions, or explaining why a seemingly precise result does not answer the question everyone thought it did.
I am particularly useful when the problem crosses boundaries. A forecasting problem can become a data-engineering problem. A software system can become a governance problem. A policy question can collapse into an argument about assumptions. This is normal.
What I Can Help With
Analysis and Decision Support
Statistical modeling and inference; forecasting and simulation; risk and uncertainty; operations research; program evaluation; benefit-cost and cost-effectiveness analysis; quantitative policy analysis; and decision support.
Artificial Intelligence and Analytical Systems
Applied artificial intelligence and machine learning; model evaluation; model-management and model-serving workflows; analytical system architecture; and the less glamorous work required to turn a model into something an organization can actually use.
Software and Scientific Computing
Research software, analytical tooling, data processing, prototypes, numerical methods, and the occasional program that exists because I needed it and it wasn’t there.
Independent Technical Review
Sometimes the useful answer is not another model. I also review analyses, assumptions, methods, code, and technical claims, especially where the result will be used to make an operational, policy, or technical decision.
What the Work Looks Like
Most consulting problems start less neatly than the headings above. The data are incomplete. The question is slightly wrong. The system already exists. Several people mean different things by the same word. A useful engagement starts by figuring out what decision actually has to be made, then choosing enough mathematics, software, or analysis to support it.
The result might be code, a model, a technical review, a decision memo, a prototype, or simply a defensible answer to a question that was harder than it first appeared.
The Long Version
If you are trying to decide whether I know anything, the rest of this site is more useful than a paragraph of adjectives. Subjects follows the ideas across the site; Scholarship has the research record; Software has working code; Writing shows how I think through problems; Books contains the things that got entirely out of hand; and Teaching shows how I explain the machinery when somebody has to learn it.
Work With Me
Have a problem? Tell me what you are trying to do.
If I can help, we can talk about what that would look like. If I cannot, I would rather tell you that than invent a reason I should.