Provider comparator, embedded-lending unit economics, BIN sponsorship, programme agreements, wallet float, VRP mandates, AP2 BaaS infrastructure mandates. Six scenarios that walk a BaaS adopter from "which provider" all the way to "what mandate does the agent runtime read." ~$40B → ~$75B addressable (Grand View Research; estimates vary by scope), every output mandate-replayable.
For teams working with Unit · Synctera · Treasury Prime · Marqeta BaaS · Galileo · useful to any product team evaluating BaaS for a programme
Start with provider economics. Model the embedded-lending product. Apply to a specific cohort (the SME passport scenario is the only one already live). Layer in the float revenue. Design the VRP sweep logic. Make the whole thing readable by an agent runtime via an AP2 BaaS mandate.
Compare BaaS provider unit economics — sponsor-bank costs, programme agreement terms, BIN sponsorship, regulatory passthrough. Surfaces the cost of choosing the wrong provider before the contract is signed.
Unit economics for embedded-lending products — origination cost, default curves, capital cost, partner take-rate. The model the product team runs before the BaaS provider sees a term sheet.
Board-ready SME readiness scorecard for Brazil + Nigeria — credit score, CCC gap, factoring + RBF blend, AP2 credit-mandate export with IFC / BCBS alignment. The only one already shipping.
Allocation mix (sweep / MMF / non-yielding) × rate scenario × cost stack → net yield in bps + annual revenue + revenue per active user. FDIC pass-through aware, with the Synapse-Evolve programme-failure cost framed as value-at-risk. AP2 treasury mandate output.
Sweep window × amount caps × idempotency × beneficiary scope across UK CMA9 / EU PSD3 / SEPA VRP / US OBA. Live 24h permitted-window timeline visualisation, allowed-beneficiary whitelist with editable pills, Policy Mandate the bank-side validator can re-read on every PISP call.
Programme-policy slots + sponsor-bank gates + BIN-tier ceilings + rail caps + handoff cadence → machine-readable Policy Mandate. Same surface used by the Agentic Runtime hub for the "Apply (BaaS)" step.
Sponsor-bank passthrough, programme agreement terms, BIN sponsorship tiers, float revenue under a given rate path, BaaS comparator economics — these all collapse to the same calc. Every BaaS evaluation rebuilds the spreadsheet from scratch. Every embedded-lending product team relearns origination economics. Every programme manager renegotiates VRP mandate parameters that should be the AP2 standard.
The catalog's BaaS cluster is the curated answer to that — modeled in the browser, exported as an AP2 programme-policy mandate the runtime can enforce. Six scenarios walk the buyer from "which provider" to "what mandate does my agent runtime read at execution time" — the last cell in the spreadsheet that always gets left out.
The cross-link scenario (24) is the same surface the Agentic Runtime hub exposes — because the BaaS team's policy and the agent-runtime team's policy converge at this mandate format.
The catalog's BaaS cluster (T152–T164) plus the rules-based engines (RBE-08, RBE-13). AP2-exportable, MCP-discoverable, none stateful.
The Tool Chain Composer renders the AINumbers catalog as a DAG with the 45 scenarios highlighted. The BaaS cluster connects upward to the agentic runtime and sideways to processors — same mandate schema across all.
Sponsor-bank economics, unit economics, programme policy, card and lending stacks — we've modeled and built them. Let's pressure-test yours before it reaches committee.
Talk to our team →