About
A practice built around one conviction.
In financial services, an AI workflow is only useful if it can be explained, repeated and evidenced. Everything else is a demo.
ALODEEL is led by Nicholas B. Fok, a strategic investment leader, product executive and company builder with two decades across institutional asset management, applied AI, and early-stage venture creation. He serves as Head of Product Solutions at OWL AI, an AI-native investment research platform for institutional firms, and as CTO of Incubator School, a global K–12 platform.
Prior to founding ALODEEL, he was Director of Global Portfolio Strategy at Liberty Mutual Investments ($80B general account), Associate Portfolio Manager at CPPIB in fixed income and currencies, and spent a decade across long/short equity (FrontPoint Partners / Copia Capital), fund-of-hedge-funds (Mesirow Advanced Strategies), and as Founder & CEO of Duli Capital, a profitable operating company.
Ownership of decision quality under uncertainty.
ALODEEL was founded after watching the same pattern repeat across financial institutions: capable teams building genuinely useful AI prototypes, then losing them at the risk review because nobody could show how an output was produced or prove it would be produced the same way tomorrow. The gap was never model capability — it was engineering discipline: versioned prompts, pinned models, typed tool contracts, validation, evaluation sets, and a retained record of every run.
So the practice was built the other way around. We start from the control environment and the review path, then design the workflow that fits inside it. That constraint turns out to produce better systems, not weaker ones: narrower model responsibilities, deterministic computation where determinism belongs, and clear escalation when the system is unsure.
We work with banks, credit unions, private equity and venture firms, private credit managers, hedge funds and fintechs. We are deliberately small, deliberately vendor-neutral, and deliberately blunt about which of your processes are not worth automating yet.
Career
Two decades across capital, AI, and company building.
- 2025 — PresentOWL AIHead of Product Solutions
Architect of an AI-native investment research platform serving institutional firms. Built the verification layer before the chat interface — grounding first, answers second.
- 2025 — PresentIncubator SchoolCTO
Co-founded a global K–12 online school. Built an AI-native platform — CRM, LMS, tutor, grading, analytics — from blank repo to live classrooms across three continents in two months.
- 2020 — 2025Duli Capital, LLCFounder & CEO
Multimillion-dollar revenue within 24 months; profitable from month one; ~200% investor IRR. Inventory run like a portfolio — position sizing, drawdown limits, kill criteria.
- 2014 — 2021Liberty Mutual InvestmentsDirector — Global Portfolio Strategy
Multi-asset factor-based SAA on $80B general account; ~$1B/yr return improvement; established the Transaction Review Committee. Deployed $1B in eleven trading days in March 2020 from a playbook war-gamed fourteen months earlier.
- 2011 — 2014CPPIBAssociate Portfolio Manager — Fixed Income & Currencies (GTAA)
$240M U.S. yield-curve portfolio at IR ~2.0; redesigned $5B fixed income book to 2× expected IR inside an active-risk budget.
- 2007 — 2011FrontPoint / Copia CapitalAnalyst — Long/Short Equity
Equity factor models, regime-based currency strategy with three consecutive positive years at ~30% returns, and $100M+ alpha from beta timing through the financial crisis.
- 2006 — 2007Mesirow Advanced StrategiesQuantitative Analyst — Fund of Hedge Funds
First encounter with the problem that defines the work today: scoring a decision when outcomes are noisy and the sample is small. Alpha style models with 15% portfolio performance improvement.
Philosophy
How we make decisions.
Say what the system cannot do
Every workflow has a boundary. We document it up front rather than discovering it in production with your clients watching.
Measure before and after
If the current process was never instrumented, we instrument it. Value claims without a baseline are marketing.
Design for the reviewer
The person who has to approve, defend or examine the output is a first-class user of the system, not an afterthought.
Leave the client self-sufficient
Documentation, training and code handover are part of the work. A dependency is not a business model we want.
Thought leadership
Credibility grounded in doing — not commentary.
Recognized at the intersection of institutional investing, entrepreneurship, and applied AI.
- FINOVATE · CDAOKeynote
Portfolio construction, governance-safe AI, and operating leverage
- Wharton Startup Challenge 2026Judge
Evaluating next-generation entrepreneurs
- Founder InstituteMentor
Vision to capital-efficient execution
- SEED SPOTMentor
Metrics and operating systems that scale
- AI & education startupsStrategic Advisor / Board
Growth, governance, and AI strategy
Next step
Talk to the person doing the work
Discovery calls are taken by the founder, not by a sales team.
