Sanat Dhir
Engineering Leader. AI Transformation. Columbia MBA.
I build the systems and teams that turn AI from a demo into a production capability — with structure, governance, and measurable ROI.
Two decades across JPMorgan, Citi, Acorns, and AI startups. Columbia MBA. Building and writing publicly about what actually works.
About
I've spent 20+ years at the intersection of technology and organizations — building systems, leading teams, and figuring out why smart companies struggle to turn good technology into real outcomes.
I've been inside that transition twice — cloud at scale in financial services, and now AI. Both times the lesson was the same: the companies that win aren't the ones with the best technology. They're the ones that redesign how decisions get made, how teams operate, and how technology connects to business results.
Technical Depth
I go deep enough to make architecture decisions, evaluate tradeoffs, and know when something is wrong before it ships. Not to do IC work — to lead credibly.
In practice:
I’ve built production RAG pipelines with evaluation harnesses, designed multi-agent development systems that outperform single-agent workflows 3-4x, and made the call to use Random Forest over an LLM when the data said it was the better choice.
Organizational Systems
Most AI adoption fails here. Companies hand out tools and hope for productivity. I design the operating models — team structures, delivery systems, governance, evaluation discipline — that let organizations actually absorb new capabilities.
In practice:
I’ve built engineering organizations from scratch, designed evaluation frameworks that catch AI failures before they ship, and created governance architectures for AI in regulated financial services — with structured workflows, refusal policies, and adversarial testing.
Business Alignment
Technology that doesn’t connect to business outcomes is a science project. I anchor everything to measurable results — whether that’s reducing client onboarding from 6-9 months to predictable timelines, cutting infrastructure costs 40%, or tying AI features to engagement metrics instead of vanity demos.
In practice:
I’ve owned multi-million dollar cloud budgets, led technical due diligence for an acquisition, and built API monetization infrastructure designed for the emerging agent economy.
I write about this publicly, build systems that test these ideas in production, and help organizations navigate the hard parts of AI adoption — the parts that aren't about the technology.
Experience
Founder
PayFence2025 – Present
API monetization platform for the agent economy. Built end-to-end — production system operational.
Head of Platform Engineering
AI Claims Platform2024 – 2025
Built a 25-person platform engineering organization across 4 teams. Delivered production RAG systems and multi-agent pipelines.
Senior Engineering Manager
Acorns2022 – 2023
Scaled engineering team from 8 to 22+ engineers serving 5M+ subscribers.
VP, Senior Engineering Manager
JPMorgan Chase2017 – 2022
Led ETF platform engineering — 200+ ETFs across 3 global regions.
Staff Software Engineer
Citigroup2014 – 2017
FX platform modernization and trading systems.
Various Engineering Roles
JPMorgan, ICAP, NYT, McKesson1998 – 2014
Early career across financial services, media, and healthcare technology.
Education
Columbia Business School
Executive MBA
University of Lucknow
B.S. Computer Science
How I Build
The best work I've done lives inside organizations I can't talk about. These are production systems I built independently — each one a window into how I think about architecture, tradeoffs, and shipping AI that actually works.
Writing
Thinking publicly about AI transformation, engineering leadership, and the future of work.
Get new posts in your inbox
No spam. Unsubscribe anytime.
In the Room
Speaking at Columbia Business School's Tech & Startup Conference, 2025
Presenting at Columbia Business School
Columbia MBA cohort visit with Jensen Huang at NVIDIA
Columbia MBA cohort visit, Google NYC
Columbia MBA exchange at London Business School
New York Stock Exchange