I started on the AMEX trading floor in the 90s, an undergrad intern in Susquehanna's index arbitrage group, building dynamic spreadsheets to spot mispricing between Japanese index futures and Nikkei warrants. The instinct that drives my work now was already operating then: find the gap in the system that nobody has bothered to fix.
By 1992 I was on the syndicate desk at CS First Boston, helping price $1.3 billion in emerging-market debt issues (including, at the time, the largest EM new issue by a lead manager) and co-authoring the firm's official handbook on the asset class. From there I traded sovereign Eurobonds and structured derivatives at Socimer, then ran emerging-markets fixed-income research at CIBC Oppenheimer: pitching yield strategies, building Excel-VBA pricing analytics, supporting an $8 billion annual trading volume.
At CIBC I built the desk's research website in 1996 (back when most analysts at name-brand firms weren't publishing on the web at all) and sat on the firm's Strategic Technology Committee as a research analyst. For years afterward, I'd find my watermarked visualizations showing up in colleagues' research papers, picked up and re-used because they made the analysis legible. That was the first time I learned the power of building things once and well.
What a trader needs at 9:31 AM is not what an analyst needs at 4:30 PM. The gap between the tool and the desk is where the work lives.
Most program and product managers running teams in financial services came from engineering or consulting. I came from the desk.
The pivot to technology
Two consultancies co-founded across a decade. Solutions-E for hedge funds and trading desks; Cambrian for global banks, six cities.
In the early 2000s I co-founded Solutions-E, a tech advisory practice for hedge funds and trading desks. The trading-floor instinct turned out to be transferable: the same arbitrage lens that finds mispriced bonds finds mispriced workflows. Both come down to spotting what nobody else has bothered to fix yet.
In 2007 I co-founded Cambrian Consulting with ex-Accenture partners. We ran program management for digital transformation initiatives across global banks, with distributed teams in London, Zurich, Tokyo, Dublin. Reported into CAOs and Heads of Enterprise Architecture. Clients included Bank of America/Merrill and UBS.
The decade in enterprise tech
DWS: SEC 22e4 compliance product; State Street cloud partnership (70% cost reduction). Morgan Stanley CCAR: Cross Market Risk workflow (30K risk entries, 50→31 days) plus Agile/Scrum coaching for two of the teams.
From 2010 to 2017 at Deutsche Asset Management, I built a regulatory compliance product for the SEC's new liquidity rule (22e4) and delivered a 70% cost reduction for 150 front-office users through a State Street cloud partnership, replacing several hundred reports across 700 users. I was also part of the team that delivered the Aladdin migration for portfolio construction.
From 2017 to 2020 at Morgan Stanley, I worked on the CCAR Stress Testing program. My scope was within Cross Market Risk: building a workflow product that consolidated 30,000 previously disparate risk entries into a single auditable seven-step workflow. That reduced Scenario Analytics cycle time from 50 days to 31 (a 39% time savings) by accelerating SME challenge turnaround. I also coached two of the teams through their transformation to Agile, specifically the Scrum framework.
AI strategy at FactSet
Four years owning AI for private markets data. Coverage: 250K → 3.7M entities. BERT accuracy: 20% → 80%. Org: ~12 across NYC, London, Hyderabad.
From 2020 to 2024 I led AI strategy and product management for the Entity Intelligence platform: private markets data solutions serving investment banking and private equity clients. Managed an org of around 12 across NYC, London, and Hyderabad through four direct reports.
Expanded covered entities four-fold (250K to 3.7M) through deployment of data extraction pipelines. Drove a 40% increase in client usage as data depth and breadth improved. Lifted BERT extraction accuracy from 20% to 80% through fine-tuning and human-in-the-loop training data workflows. Cut processing time 90% with a tiered data refresh architecture.
The work taught me the difference between a model that works in a notebook and one that holds up in a live institutional environment.
What I'm doing now
I'm currently engaged at a leading global PE firm, building a platform designed to save senior tech leaders hundreds of hours.
Stanford GSB Harnessing AI (2022), MIT Applied GenAI (2023), and Stanford continuing studies coursework on LLMs round out a deliberate three-year toolkit rebuild.
What I'm thinking about
Agentic AI patterns for institutional workflows. Advisor-facing AI in private wealth. How legacy tech orgs scale generative AI responsibly: the data discipline that has to come before the AI tooling, not after. The space between the model and the desk where program management actually earns its keep.
Education: MIT, BS Operations Research and Management Science (US Navy ROTC, served on the USS Lang FF-1060). Columbia, MS Technology Management. Fluent Spanish.
Based in Manhattan. I work fully on-site, by preference, not circumstance. Second home in Park City, Utah, where the Alta name comes from.