Twenty years shipping AI inside regulated finance.
Khurram Tehseen has spent his career building the thing most of the industry only talks about: AI systems that make regulated financial decisions in production — in front of auditors, examiners, and committees paid to be skeptical.
20 years in regulated financial services · 160-person global organization · FSA, FCIA · Claude Certified Architect
01Most recently
As Managing Director of Data & AI at an $8B+ alternative asset manager, I served as de facto Chief AI Officer reporting to the CEO, leading a 160-person global organization. The platforms my teams shipped generated $2.2B+ in investor profits at a 14% IRR.
Four systems account for most of that: an LLM underwriting engine running in a HIPAA- and SEC-regulated environment (33% faster at 99.9% decision accuracy); an agentic monitoring workflow that took a core process from 26 days to three; a platform modernization that cut valuation runtimes 97%; and underneath it all, a proprietary data asset spanning 1.5 billion claims records.
The order matters. The data platform came first — the headline numbers were only possible because someone did the unglamorous work of resolving entities and fixing pipelines. Which is why, when a client asks me to start with the use-case list, I usually ask to see the data layer instead.
02Before that
At Manulife, I launched the firm’s first global advanced-analytics organization — 160+ professionals across three continents, built from a standing start. The work improved underwriting and reserve accuracy by $754 million: a balance-sheet number, signed off by the people whose job is to find the flaw in it.
Track record at a glance
| $2.2B+ | Investor profits from production AI platforms, at a 14% IRR |
| $754M | Underwriting and reserve accuracy improvement, global insurer |
| 99.9% | Decision accuracy, production underwriting engine (33% faster) |
| 26 → 3 days | Core monitoring process, rebuilt as an agentic workflow |
| −97% / −35% | Valuation runtime and infrastructure cost, post-modernization |
| 160 | Largest organization led, across three continents |
03Why fractional
The firms that need senior AI leadership most — mid-market funds, insurers, and the platforms consolidating them — are exactly the firms that can’t justify a full-time C-suite AI hire, let alone the six-month search that precedes it.
The fractional model isn’t a compromise. It’s the correct amount of a very expensive thing.
Strategy, architecture, and governance are led by me personally. Implementation is delivered by a vetted bench of specialists — document-AI engineers, data platform builders, quantitative modelers — working under my review. Senior judgment on every decision, specialist hands on every build, no pyramid of juniors learning on your dime.
04Credentials, and when they matter
I’m a Fellow of the Society of Actuaries, a Fellow of the Canadian Institute of Actuaries, and a Claude Certified Architect. For insurance clients, the actuarial credentials are often the difference between a model that gets discussed and one that gets adopted. For everyone else, the operating record above is the credential.
05How I work
Diagnose before prescribing. Engagements start with your people and your documents, not a framework deck.
Numbers or it didn’t happen. Every recommendation carries an impact estimate, a cost band, and a time-to-value.
Governance is a chapter, not a caveat. A system that can’t survive scrutiny isn’t an asset — it’s a finding waiting to be written.
Vendor-neutral, by policy. No referral fees, no commissions. One financial interest in the recommendation: that it works.
Your artifacts, your environment. If an engagement ended tomorrow, you’d keep everything of value.
Khurram Tehseen
Founder, Operating Alpha AI · khurram@operatingalpha.ai · LinkedIn