Jonathan F. Schachter, PhD — Research Dossier for the Columbia IEOR Application
Original research by Delta Vega — a 65-program global survey of AI in financial engineering curricula — paired with a complete map of Columbia's IEOR department and the single competency gap that runs through all of it.
Original Research — Delta Vega Global Survey
Because our model-risk textbook contains an extensive chapter on AI model validation, Delta Vega conducted a structured survey of how widely AI — machine learning, deep learning, data science — is taught across master's-level financial engineering, mathematical finance, and computational finance programs. The intent: to help instructors evaluate AI's place in a curriculum, and to help students differentiate between programs.
| Country | Degree Programs |
|---|---|
| United States | 26 |
| United Kingdom | 25 |
| Australia | 4 |
| Canada | 2 |
| Singapore | 2 |
| Belgium · France · Germany | 1 each |
| New Zealand · South Africa · Sweden | 1 each |
Table 2 — Terminology in AI Course Titles
"Machine Learning" dominates the curricular vocabulary — appearing in titles six times as often as the next term. The field has a common language, but not yet a common standard for what comes after the model is built.
Table 3 — The Eight Programs That Require an AI Course (12% of cohort)
| Institution | Degree Program | AI / ML Courses Listed |
|---|---|---|
| Baruch / CUNY | MS Financial Engineering | Data Science I: Big Data in Finance (MTH 9898); Data Science II: Machine Learning (MTH 9899); NLP (MTH 9796) |
| Bayes Business School | Quantitative Finance MSc | Optional AI / ML electives |
| Birkbeck College | Quantitative Finance with Data Science MSc | Data Science track embedded |
| Boston University | MS Mathematical Finance & FinTech | Advanced ML Applications for Finance (QST MF 815); Deep Learning & Statistical Learning (QST MF 850) |
| Carnegie Mellon | MS Computational Finance | Machine Learning I; Machine Learning II (46927) — required |
| Columbia (IEOR) | MS Financial Engineering | ML for FE & OR (E4525); Deep Learning (E4742); AI Applications in Finance (E4737) — all elective |
Full dataset published to the Teaching Financial Model Risk & Governance group. The survey is an exhaustive institution-by-institution pivot; the rows above are a representative extract.
"Should AI courses be part of an FE master's program? What specifically should be taught? Should it be required or optional — and should dissertations, capstones, or independent research be required as adjuvants?"
The open questions this dossier sets out to answer for Columbia
This survey is Jonathan's credential of authority. He is not a candidate offering opinions about AI in finance — he is the author of the only global benchmark of how AI is actually taught across the field, grounded in a textbook chapter on AI model validation. Leading with original research reframes the entire application: the committee meets a scholar who has already mapped their field's curriculum landscape, and who arrives with data rather than assertions.
IEOR runs five separate MS programs, five undergraduate majors, two PhD tracks, and eight research centers. The five graduate programs are: MSFE (financial engineering), MSOR (operations research), MSIE (industrial engineering), MSE (management science, joint with CBS), and MSBA (business analytics, joint with CBS). All are STEM-certified.
His cover letter should demonstrate he understands the full department — not just MSFE. The hiring committee is IEOR faculty, not the MSFE director alone. Showing he can contribute across programs immediately widens his appeal and differentiates him from candidates who only know the finance track.
Jonathan isn't proposing to add AI to a department that ignores it. He's proposing to fill a specific, named gap in a department that's already all-in on AI — a much stronger position. He should open by acknowledging IEOR's commitment to ML/analytics, then pivot to the missing layer.
Across all six programs (MSFE, MSOR, MSIE, MSE, MSBA, undergrad), the pattern is identical:
| Program | AI in Curriculum | LLMs | Model Risk / Uncertainty / Governance |
|---|---|---|---|
| MSFE | Electives | None | Light touch |
| MSOR | Elective + concentration | None | None |
| MSIE | Elective option | None | None |
| MSE | Elective / FinTech | None | None |
| MSBA | Core required | None | None |
| Undergrad | Named AI course | None | None |
Jonathan's differentiation is not just for MSFE students — it's a department-wide gap. A course on trustworthy AI / model risk / uncertainty quantification would serve MSBA students (taught to build analytics models but not validate them), MSOR ML-concentration students, MSE FinTech students, and MSFE students simultaneously. This makes him a hire who serves the whole department.
In 2024 Columbia launched the Center for AI and Responsible Financial Innovation (CAIRFI), a five-year partnership with Capital One, with the stated mission of "the responsible advancement of AI in financial services." Its funded research spans AI agents in finance, trustworthy lending, domain-specialized LLMs, AI privacy and security, and causal inference — the precise problem set Jonathan's model-risk and governance work addresses.
CAIRFI demonstrates that responsible AI is already a strategic priority at Columbia — on the research side. Jonathan's contribution is its pedagogical counterpart: a teaching program that trains students in the validation and governance the center researches. Framing his proposal as the classroom complement to an existing institutional commitment signals genuine familiarity with the department and positions him as a colleague who extends its agenda, not one arriving with an unrelated idea.
Professor Henry Lam works in risk analysis, simulation, and uncertainty quantification — the closest existing faculty research to Jonathan's area. His emphasis, however, is UQ for operations research and simulation, rather than AI/ML governance in finance under the current regulatory regime.
Jonathan should frame his work as complementary, not competing. Lam brings uncertainty quantification to classical simulation models; Jonathan brings it to machine-learning and AI systems under SR 26-2. Same rigour, adjacent domains — a natural basis for collaboration. Acknowledging Lam directly shows the committee Jonathan has read the faculty and understands precisely where he adds something new.
MSIE's Healthcare Management concentration and MSBA's Healthcare Analytics concentration are substantial programs. Clinical AI — diagnostic models, readmission prediction, clinical decision support — carries model-risk consequences arguably greater than finance. No course anywhere in IEOR currently addresses AI governance or uncertainty in healthcare.
Jonathan need not lead with healthcare — it isn't his primary domain — but it strengthens the case: model risk and uncertainty quantification are as critical for clinical AI as for financial AI, and IEOR teaches both. It demonstrates that an investment in him serves multiple programs across the department.
The Bottom Line
Jonathan arrives with original research, a model-risk textbook, regulatory fluency under SR 26-2, and a teaching program that serves as the practical complement to what CAIRFI already funds on the research side.