Fintech: how the sector works
how fintech works: unbundling and re-bundling financial services, the main models (payments, lending, neobanks, embedded finance), and the regulatory/licensing reality.
Fintech reshapes how money moves, is stored, lent, and invested by unbundling traditional banking into specialized digital services. This block gives you the operating map: how payments, lending, banking-as-a-service, and wealthtech actually function end to end, who captures value across issuers, processors, networks, and platforms, and how incumbents, challengers, and Big Tech compete for distribution and margin. You will learn the regulatory architecture (licensing, AML/KYC, data and payments rules) that shapes what is legally possible, plus the market sizing, benchmarks, and vocabulary needed to read the sector fluently. Together these modules build the baseline literacy required before going deeper into specific fintech verticals or deals.
What you'll master
- Map a fintech value chain end to end and identify where margin and value actually concentrate
- Assess competitive position of an incumbent, challenger, or platform player given power dynamics in the chain
- Identify which regulations and licenses apply to a given fintech business model and what compliance they require
- Use market sizing figures, benchmarks, and standard calculations to evaluate or pitch a fintech opportunity
Key terms
Modules
How fintechs break apart and reassemble the banking value chain across four core business models.
Who competes, who controls the customer, and how power and margin shift across the fintech ecosystem.
The major laws, compliance duties, and regulatory bodies shaping fintech operations.
Market sizing, acronyms, benchmarks, and quick calculations every fintech professional should know.
Latest articles
Recent articles from the blog that apply to Fintech.
- AISR 11-7 still bites, and gradient boosting just made the wound worseSR 11-7 was written for logistic regression, but banks are now deploying gradient boosting, neural networks, and foundation model-powered scoring into production. This piece unpacks what explainability actually means under the guidance, why examiners are pushing harder on it in 2026, and where the governance frameworks genuinely break down.
- Finance$800mn at an $8bn floor: what Airtel Money's London IPO demands from an African fintech CFOAirtel Money is preparing to file prospectus documents for what could be one of London's largest listings in recent years, targeting $800mn in proceeds at a valuation of $8bn to $9bn. The preparation required to reach that point tells CFOs more about IPO readiness than any generic checklist.
- DataThree pipeline design decisions that determine whether your AML model survives its first regulatory examinationMost fintech fraud and KYC/AML pipelines fail not because the models are weak but because the data architecture cannot defend itself under examination. This playbook walks through the design sequence that keeps you compliant, explainable, and operationally credible when regulators arrive.
- FinanceRevolut's dual listing play and what it signals for fintech CFOsRevolut is preparing to list simultaneously in New York and London, a structural choice that reveals as much about equity story architecture as it does about exchange selection. For CFOs in high-growth fintech, the decisions behind that choice are worth studying carefully.
- AIBuilding credit decisioning models that survive fair-lending scrutinyAI-driven credit models can cut decisioning time and expand credit access, but a single fair-lending violation can trigger enforcement actions that dwarf any efficiency gain. This playbook shows banking AI leaders how to build, document, and defend models that hold up when the OCC, CFPB, or DOJ come knocking.
- DataHow Tala built a credit engine for the world's most invisible borrowersTala lends to borrowers who don't exist in any credit bureau, using smartphone data as a substitute for a credit file. Here is what their model actually does, what it has produced, and what fintech data leaders can reasonably take from it.