01Overview
Transforming receivables into capital.
Businesses get paid early.
Institutions earn short-duration yield.
MINT allows massive computation and machine learning to operate continuously across commercial receivables.
Each claim remains in a current, auditable state for eligibility, pricing, routing, and monitoring.
02Shift
Infrastructure > Agents
Great infrastructure compounds as models advance.
Let the machines run.
Human-designed rules, features, and task-specific agents deliver fast early gains.
They also limit the system to what people already know.
The largest advances come from general methods that use massive computation to search and learn beyond those assumptions.
03Problem
Liquidity Gridlock
Legacy financial architecture traps valuable claims and blocks liquidity.
Institutional capital cannot finance what it cannot discover, verify, price, and monitor.
Commercial evidence is fragmented across orders, performance, acceptance, invoices, deductions, remittances, and payments.
No single document contains the complete current claim.
The result is a static record instead of a maintained claim state.
04Solution
Capital Infrastructure for Machines
Machine speed liquidity for small businesses.
A continuously underwritten asset class for institutional capital.
MINT turns receivables into programmable collateral.
MINT captures fragmented commercial evidence and maintains a current claim state.
Models use that state for eligibility, pricing, routing, and monitoring.
05Opportunity
Hidden Corporate Credit
Make commercial claims visible, verifiable, measurable, and ownable.
Connect financeable claims to institutional capital.
Small-business commercial claims remain largely outside institutional credit markets.
The obligations already exist. The capital already exists.
The opportunity is the institutional market between them.
06Properties
Control and Compression
Control makes collateral verifiable.
Compression makes it programmable.
Control keeps every claim state current and preserves its audit trail.
Compression converts fragmented commercial evidence into compact, comparable claim states.
Together, they power Search & Learn.
Capture · Rationalize · Standardize
Capture
Fragmented commercial evidence becomes knowable.
Rationalize
Events resolve into a coherent current claim state.
Standardize
Claim states become comparable and programmable.
MINT Receivables
ControlCompression
07Control
Discovery Through Compute
Every controlled claim enters a common state space.
Compute discovers financeable structures no fixed rule set could predefine.
Control and Compression convert fragmented commercial evidence into canonical claim states.
Search runs across claims, events, counterparties, and time rather than one transaction at a time.
Recurring state patterns become candidate underwriting logic, asset structures, and cohorts.
08Compression
Continuous Underwriting
Each resolved claim produces an outcome signal.
Underwriting compounds at network scale.
Models compare predicted claim states with realized cash, deductions, disputes, recoveries, and timing.
Learning identifies which evidence patterns and state transitions predict value, risk, and liquidity.
Each learning cycle improves the next search.
09Reconstruction
Financeable Net Claim
Finance the debt that survives.
The value was computationally invisible, not absent.
MINT separates the stated claim into supported principal, known adjustments, and conditional amounts.
Reserves and unresolved residual remain visible.
As evidence arrives, the financeable net claim changes.
10Asset
Programmable Capital
A standardized claim institutional capital can price and own.
Built on familiar debt. Purpose-built for speed, transparency, and defined risk.
The asset is a continuously reconstructed net cash claim.
Evidence, eligibility, adjustments, reserves, and payment state remain current through settlement.
Institutions buy the claim while businesses unlock liquidity without financing the company.
11Cohorts
Repeatable Supply
Standardization makes claims comparable, not identical.
Recurring patterns reveal a cohort.
Repeated qualified fit creates institutional supply.
MINT searches across recurring commercial conditions, current claim states, and realized outcomes.
Claims with the same stable pattern form a candidate cohort.
Institutional Buy Boxes determine which current claims qualify for purchase.
EXPLORE COHORT EXAMPLESCLOSE COHORT EXAMPLES
Recurring conditions × current claim state × observed outcomes → candidate cohort
- 01Return-window survivorsReturn-resolved receivable
- 02Late-delivery claimsPenalty-bounded receivable
- 03Partially fulfilled POsLine-item microreceivables
- 04Delivered, awaiting acceptanceEvent-vested receivable
12Capital
Dedicated Capacity
Repeated qualified fit creates a capital lane.
Institutional demand is defined before claims are purchased.
Each institution defines evidence, credit, tenor, concentration, exception, and economic requirements.
MINT routes matching claims into dedicated capacity as qualified supply forms.
The institution retains mandate and purchase authority.
13Scale
Compounding Infrastructure
More claims expand the search space.
More outcomes improve underwriting.
Each claim adds evidence, state transitions, and realized performance.
Each institutional mandate adds another definition of financeable fit.
The same infrastructure compounds across claims, buyers, and time.