Parlay

The removal test leaves a static intake form, which is precisely what the company defines itself against, distinguishing its layer from conventional origination systems on the grounds that it integrates real-time borrower insights, behavioural analytics and predictive modelling rather than holding fixed rules.

Applicant profiles update dynamically as data arrives, credit box validation applies alternative financial data and business performance metrics alongside programme rules, and behavioural tracking drives re-engagement of applicants who stall.

The product is positioned before the decision rather than at it, qualifying and packaging applicants so that underwriting receives better-prepared files, and the company is explicit that lenders maintain rigorous underwriting standards while gaining efficiency. That placement is itself an oversight choice, since the consequential judgement remains with the credit function.

Held at B because filtering at the top of the funnel is consequential in its own right, and no threshold, override or review mechanism is described for applicants the system screens out before a human sees them.

No accuracy, validation, error rate or evaluation method is published for models that score applicants on alternative data and determine which reach underwriting. Dynamic credit scoring incorporating cash flow, business health and alternative financial data is described in capability terms without any account of how it is built, tested or monitored, which is the disclosure a bank's model risk function would need before allowing it to filter its pipeline.

Two community lenders are named, though by an independent guide rather than by the company, and partnerships exist with a major card network's small business arm and with the venture firm that led its two million dollar seed, itself a joint venture bringing together bank investors. One customer testimonial cites reduced manual work across teams, with the percentage garbled in the source. No deployment count, loan volume or approval outcome is published, and the funding round is small, so the record is early rather than substantial.

No boundary statement was located. The platform observes application behaviour, qualification outcomes and drop-off patterns across community lenders competing for small business customers in overlapping markets, which is exactly the material that would sharpen its qualification models. Nothing states whether that learning is pooled or isolated per lender.

No data protection agreement, retention schedule, subprocessor list or consent framework was located, and two collection mechanisms deserve naming. The platform automatically gathers financial, credit, industry and tax data through pre-configured interfaces, and it applies real-time behavioural tracking to applicants, detecting where they hesitate or abandon the process. Behavioural observation of someone applying for credit is a different category of data from what they submit, and nothing describes whether applicants are told it occurs.

No attestation, certification, trust centre or enumerated framework was located. The platform ingests tax records and financial data directly from applicants and connects into lenders' origination systems, and no control documentation is published for an institution beginning its assessment.

A specific federal lending programme is the operating frame throughout rather than an afterthought, with eligibility verification, compliance reporting and applicant scoring all described as aligned to that programme's rules, and the company notes explicitly that scoring must adapt as those guidelines evolve. Compliance reports are pulled in real time as part of eligibility checks. Held at B because no regulator, statute or fair lending requirement outside the programme is named, and pre-decision screening has its own regulatory implications that are not addressed.

The access argument is specific about who is being missed and why: many small business applicants fail rigid criteria because of outdated credit models, producing high rejection rates, and enriched profiles incorporating alternative data are intended to surface strong applicants who do not meet traditional credit score cut-offs.

What distinguishes this from similar claims elsewhere is that the company also acts on the near-miss population directly, identifying and nurturing applicants who are close to qualifying rather than discarding them. Held at B because no fairness testing, approval rate analysis by borrower group, or measured outcome accompanies any of it.

This is the only vendor in the index that gives the rejected applicant somewhere to go. Where every comparable platform's declined borrower simply disappears, this one identifies applicants close to qualifying and guides them to strengthen their financials before reapplying, and detects abandonment points in order to remove the barrier rather than record the loss. That is recourse built into the product rather than promised in terms.

Held at B because it is not liability: no guarantee, indemnity or correction process exists, an applicant cannot see or contest the qualification logic, and the guidance is offered at the company's discretion rather than as a right.

No base model, provider, hosting arrangement or subprocessor is identified, and the alternative data underpinning enriched profiles and credit box validation is described only by category, covering financial, credit, industry and tax sources without naming any provider. For a scoring layer that determines which applicants advance, the provenance of those inputs is what determines who is filtered out.

Integration is the stated architectural principle, described as interface-forward and designed to connect with existing origination platforms specifically to reduce technology complexity for institutions with small teams, and pre-configured connections pull borrower financial records, tax data and programme compliance reports in real time. That addresses the practical barrier for a community bank without integration capacity. Held at B because no origination system, data provider or core platform is named individually.

No hosting provider, region selection, residency commitment or private deployment option was located. Community banks handling small business tax records and financial statements would raise this during vendor review, and nothing is published.

No pricing, packaging or basis of charge was located. The value proposition is explicitly about making small loans profitable to originate, which makes the platform's own cost per application directly material to whether the argument holds, and none of it is published.

The buyer is defined tightly and deliberately as community banks and credit unions serving small business customers, and product coverage within that is broad, described as a single intake system spanning working capital, government-guaranteed, acquisition, small scored and commercial and industrial lending. Held at B because the market is one country and one segment, and no institution size range or deployment breadth is evidenced.