New York · Enterprise AI

Enterprise AI built for New York's regulators, not just its demos

New York's banks, insurers, law firms and hospitals face some of the most specific AI and cybersecurity rules in the country. I build agents, retrieval systems and private AI that move a real KPI and hold up in front of a regulator. Fixed scope, fixed price, weekly demos.

Talk about your New York team Industries I work with
10.02MNonfarm jobs in the New York-Newark-Jersey City metro, 2025 annual average Source
+7.9%Growth in New York metro real GDP, 2018 to 2023 (computed from the chart data below) Source
20.11MResidents of the New York-Newark-Jersey City metro, 2025 Source

The New York metro is one of the largest economies in the country and one of the densest markets for advanced professional services anywhere: banking, insurance, law, accounting and management consulting, stacked on top of each other in Manhattan and spread across New Jersey and Long Island. JPMorgan Chase, Citigroup, Goldman Sachs and Morgan Stanley are headquartered here, alongside insurers such as MetLife, AIG, New York Life and TIAA, and the NYSE and Nasdaq.

That density cuts both ways for AI. The upside is enormous volumes of text-heavy, repetitive, high-value work. The catch is that New York regulators got specific early: the Department of Financial Services has had a cybersecurity regulation for financial companies since 2017, and New York City requires bias audits for automated hiring tools. I design AI systems that assume that scrutiny from the start, and I tie each one to a KPI the business already tracks, the same way I managed more than $50M in ad spend over a 15-plus year career in marketing and data analytics.

New York's industries, and where AI moves the number.

The sectors that anchor the New York economy, and the specific work I do for each.
01 / Industry

Banking and capital markets

Wall Street is still the center of gravity. JPMorgan Chase, Citigroup, Goldman Sachs and Morgan Stanley are headquartered in the city, and every regulated financial company in the state answers to NYDFS and its Part 500 cybersecurity rules. In May 2026 DFS issued an industry letter on heightened cybersecurity risks from frontier AI models.

How I help

I build agents for KYC and onboarding reviews, trade and payment exception handling, and research and policy Q&A with citations, all deployed with access controls, logging and model risk documentation that fit a Part 500 program. The KPIs are review hours per file, exception aging and the percentage of outputs accepted by the second line without rework.

Enterprise AI for Financial Services
02 / Industry

Insurance

MetLife, AIG, New York Life and TIAA all run their headquarters from New York, which makes the city a hub for life, annuity, retirement and commercial lines, each with long policy documents and heavy servicing volume.

How I help

I build systems that read policy contracts, applications and correspondence, extract the fields that matter and route service requests with a drafted response for the agent to approve. For underwriting, retrieval over guidelines and prior cases shortens review. I measure handle time per request, straight-through processing rate and quality-review pass rates.

Enterprise AI for Insurance & Underwriting
03 / Industry

Law firms and corporate legal

New York has one of the highest concentrations of law, accountancy, banking and consulting firms of any city in the world, and corporate legal departments here manage enormous contract and discovery volumes.

How I help

I build private RAG and knowledge systems over a firm's own precedents, clause libraries and matter files, plus contract review agents that flag deviations from playbook positions. Everything stays inside the firm's environment with matter-level permissions. The KPIs are associate hours per first-pass review, turnaround on standard agreements and citation accuracy on internal research answers.

Enterprise AI for Legal & Law Firms
04 / Industry

Health systems

Health care employs roughly 565,000 people in New York City across more than 70 hospitals. NYC Health + Hospitals describes itself as the nation's largest public health care system, alongside academic systems such as Mount Sinai.

How I help

I focus on the administrative side: referral and intake processing, prior authorization packets, denials follow-up and staff questions against internal policies, running on private infrastructure with PHI controls and human sign-off. The KPIs are days in accounts receivable, denial overturn rates and staff minutes per case.

HIPAA-Compliant Enterprise AI for Healthcare

New York by the numbers.

Public data on the size and direction of the New York economy.
New York metro real GDPReal GDP of the New York-Newark-Jersey City MSA, billions of chained 2017 dollars, 2018 to 2023
05001,0001,5002,0002018201920202021202220232018: $1,766.3B2019: $1,801.1B2020: $1,744.7B2021: $1,834.5B2022: $1,875.1B2023: $1,905.2B$1,766.3B$1,905.2B
Show the numbers
Year$B
20181,766.3
20191,801.1
20201,744.7
20211,834.5
20221,875.1
20231,905.2

Source: U.S. Bureau of Economic Analysis via FRED. FRED series RGMP35620 (millions of chained 2017 dollars, New York-Newark-Jersey City, NY-NJ-PA MSA), converted to billions and rounded to one decimal. FRED marks this series as discontinued; 2023 is the latest value published.

New York metro nonfarm jobsTotal nonfarm employment in the New York-Newark-Jersey City MSA, annual average, thousands, 2020 to 2025
8,5009,0009,50010,00010,5002020202120222023202420252020: 8,770.4K jobs2021: 9,073.3K jobs2022: 9,606.6K jobs2023: 9,809.5K jobs2024: 9,967.1K jobs2025: 10,024.7K jobs8,770.4K jobs10,024.7K jobs
Show the numbers
YearK jobs
20208,770.4
20219,073.3
20229,606.6
20239,809.5
20249,967.1
202510,024.7

Source: U.S. Bureau of Labor Statistics via FRED. MSA-level BLS CES series SMU36356200000000001A, annual frequency (BLS annual average, not seasonally adjusted), thousands of jobs. Values as published.

New York's AI ecosystem is built around its universities and its buyers. Cornell Tech, the applied sciences graduate campus Cornell built with the Technion on Roosevelt Island, sits alongside Columbia and NYU in the city. Those three, plus CUNY, SUNY, the Icahn School of Medicine at Mount Sinai and the Flatiron Institute, are among the members of Empire AI, the state-backed consortium that launched the first phase of its shared AI computing system at the University at Buffalo in October 2024. The other half of the ecosystem is demand: many of the fintech and enterprise software companies of Silicon Alley sell into the banks, insurers, media companies and law firms headquartered a few blocks away. For an enterprise, the result is plenty of vendors and plenty of pilots. The scarce part is getting one of them into production against a number.

I am based in Minnesota and work remote-first with New York teams. That fits how most New York organizations already run: distributed across Manhattan, New Jersey and Long Island, with stakeholders who live on video calls. Each engagement has a fixed scope and price, a KPI agreed on day one and a working demo every week. I come on-site for the moments that benefit from a room: the kickoff workshop, sessions with risk, compliance or general counsel before anything touches production, and launch. When a model risk or security review is part of your process, I plan the documentation for it into the scope rather than treating it as an afterthought.

Questions from New York teams.

We are regulated by NYDFS. How do you handle Part 500 and the AI guidance?

I treat the AI system as part of your cybersecurity program from the start: access controls, data classification, logging, vendor risk review for any model provider, and a clear record of what data the model can see. Where the risk calls for it I deploy models privately so sensitive data never leaves your environment. I am not your compliance counsel, but I build so your compliance team can sign off.

Can you build AI for recruiting without running into Local Law 144?

If a tool substantially assists hiring or promotion decisions for New York City roles, it needs an annual bias audit, public disclosure of the results and advance notice to candidates. I scope HR use cases with that in mind, either keeping the AI out of the decision or designing the logging and data needed for an independent audit.

Our law firm is worried about client confidentiality with generative AI. What does a safe setup look like?

Private deployment inside your environment, matter-level permissions that mirror your document management system, no training on client data, and answers that cite the source documents so lawyers can verify every claim. We start with internal knowledge and precedent search before anything client-facing.

We have run a dozen AI pilots. Why would this one reach production?

Because it starts with the metric, not the model. I pick one workflow, baseline its current cost and cycle time, and build weekly toward that number with the owners of the process in the room. If the number does not move, we find out in weeks, at a fixed price.

Put AI to work in New York, tied to a number.

Tell me the process or the problem. You'll get a straight answer on what it's worth and what it would take.

Let's talk

Other markets I work in.

Sources

  1. FRED: Real GDP for New York-Newark-Jersey City, NY-NJ-PA (MSA), RGMP35620
  2. FRED: All Employees, Total Nonfarm in New York-Newark-Jersey City, NY-NJ-PA (MSA), SMU36356200000000001A
  3. FRED: Resident Population in New York-Newark-Jersey City, NY-NJ (MSA), NYTPOP
  4. Wikipedia: Economy of New York City
  5. NY Department of Financial Services: Cybersecurity Resource Center
  6. NYC Department of Consumer and Worker Protection: Automated Employment Decision Tools
  7. Empire AI: Who We Are
  8. NYC Health + Hospitals: Nation's Largest Public Health Care System Launches IT Upgrade