Analytics & AI leadership
Translate organizational priorities into defined analytical work, clear ownership, and a practical path to delivery.
CHICAGO · ANALYTICS · APPLIED AI
Analytics and AI leadership for complex decisions.
I lead analytical teams and technical programs that connect applied statistics and machine learning to operational priorities.
View selected workFrom uncertainty to an informed decision.01 — SELECTED RESEARCH
HOW I WORK
Depth in the analysis.
Clarity in the decision.
Statistical modeling
Make assumptions explicit, compare strong baselines, and quantify the uncertainty that changes a decision.
Operational improvement
Connect model outputs to resource allocation, financial reconciliation, and the work people actually perform.
SELECTED WORK
Questions worth
getting right.
Four independent research projects connect statistical methods to concrete decisions. Starter code is available; empirical evaluations are in development.
RESEARCH FRAMEWORK01 / 04
Which feasible roster offers the best balance of expected performance, cost, and uncertainty?
- 01Forecast uncertainty
- 02Apply constraints
- 03Compare decisions
Independent portfolio research · In development
DECISION SCIENCE · BASEBALL
Resource allocation under uncertainty
How should a fixed budget be allocated when future performance is uncertain? Baseball provides a concrete setting for a broader operational problem.
Research in development
RESEARCH FRAMEWORK02 / 04
Which transactions can be matched confidently, and which need an analyst's attention?
- 01Generate candidates
- 02Resolve ambiguity
- 03Route for review
Independent portfolio research · In development
APPLIED ML · FINANCIAL OPERATIONS
Reconciliation with human review
Connect candidate transaction matches to an auditable review process. The research focuses on ambiguity, missed exceptions, and analyst workload.
Research in development
RESEARCH FRAMEWORK03 / 04
How much of an observed outcome change can be attributed to an intervention under explicit assumptions?
- 01Define intervention
- 02State assumptions
- 03Estimate effects
Independent portfolio research · In development
CAUSAL INFERENCE · EXPERIMENTATION
Measuring an intervention's effect
Separate a change in outcomes from the effect of a decision. Start with a transparent estimator and make its identifying assumptions inspectable.
Research in development
RESEARCH FRAMEWORK04 / 04
When does a probability estimate imply positive expected value at an available price?
- 01Estimate probability
- 02Observe price
- 03Evaluate uncertainty
Independent portfolio research · In development
QUANTITATIVE RESEARCH · MARKET PRICING
Probability, price, and expected value
Investigate the relationship between calibrated predictions and market prices. Keep modeled opportunity, realized performance, and market comparisons distinct.
Research in development
ABOUT MICHAEL
Technical depth.
A wider perspective.
My work spans enterprise analytics, financial reconciliation, applied machine learning, and technical program leadership. I bring a doctorate in applied multivariate statistics and experience guiding analytical work in complex organizations.
I’m interested in the point where a model becomes useful: what decision it supports, what evidence justifies it, and how its limitations affect the people using it.
Baseball offers a particularly rich setting for these questions—limited resources, uncertain performance, and outcomes that can be tested against a record.
Professional background on LinkedIn ↗LET’S CONNECT