BeginnerStart here. No prior knowledge assumed.
0 of 5 complete
Lesson 1
Why Every Fintech AI Looks the Same
Same models, same data vendors, same public rules — and where the real differentiation actually sits.
Lesson 2
The Fintech AI Landscape
The nine problems a fintech must solve, and how the modules fit together.
Lesson 3
Regulatory Spine — India
RBI, video KYC, Account Aggregator, digital lending and DPDP — what an AI system may and may not do.
Lesson 4
Regulatory Spine — Global
EU AI Act, GDPR Article 22, US fair lending and how jurisdictions differ on automated decisions.
Lesson 5
Identity and Onboarding
KYC, document pipelines, liveness and the review queue. Every regulated product starts here.
Why do most fintech AI products look alike?
What is the first thing every regulated financial product must do?
IntermediateBuild on the basics.
0 of 5 complete
Lesson 1
Credit and Underwriting
Thin-file lending, alternative data, and adverse action reasons you can defend.
Lesson 2
Fraud and Risk
Real-time scoring, velocity and graph features, mule detection and the false positive problem.
Lesson 3
AML and Compliance
Sanctions screening across transliterations, threshold analysis and FIU-IND reporting.
Lesson 4
Payments and Reconciliation
Matching cascades, exception handling, UPI dispute rails and ledger design.
Lesson 5
Customer Operations
Grounded support agents, escalation design and the recovery conduct rules.
What is the strongest alternative-data credit signal in India?
Why is step-up authentication preferred over blocking in fraud?
AdvancedDepth, edge cases and strategy.
0 of 5 complete
Lesson 1
Wealth and Advisory
The line between education and regulated advice, and suitability as deterministic rules.
Lesson 2
Infrastructure
Core banking, the thin ledger pattern, data architecture and vendor exit readiness.
Lesson 3
Governance
Model inventory, risk tiering, independent validation, bias testing and kill switches.
Lesson 4
Licensing and Access
Permission paths, capital requirements and the data you cannot buy directly.
Lesson 5
Build Playbook and Go-Live
Sequencing the whole build, and the gate every line of which must be true.
Why must a generative model never produce an investment recommendation?
Why are third-party models tiered HIGHER risk, not lower?
NextThe course explains the capabilities. These walk one product end to end.
Then build one
Every guide takes a single product through eight steps: the options at each one, what goes in and what comes out, what it costs, and what breaks. 19 products, four regulators. Start with the one closest to what you are actually building.