Credit decisioning for SME lenders
Maqina is a swarm of specialist AI agents that deliberate to a single, auditable credit decision — so you approve faster, turn away fewer good borrowers, and can explain every call you make.
The problem
Manual underwriting takes days. SME borrowers don't wait — they take the first offer that lands. Every hour is pull-through you lose.
Rigid scorecards reject solid businesses with thin files, seasonal cash flow, or patterns the template can't read. That's revenue you turn away.
Black-box outputs create audit findings and Consumer Duty exposure — risk that lands on your desk, not the vendor's.
The engine
Not one black-box model. A committee of specialist agents — each an expert in one signal — that share findings, challenge each other, and converge on a decision that shows its work.
How it works
Live queries to validated sources — company registries, bureau, bank-statement APIs. Every datum sourced and timestamped.
Specialist agents activate — cash-flow, collateral, industry-risk, behavioural, KYC.
Agents share signals, challenge each other, and converge by weighted consensus — committee review, in real time.
A decision with a confidence score, an agent-by-agent audit trail, and a regulator-ready explanation.
Why Maqina
Cash-flow and behavioural agents read seasonality and atypical patterns a rigid template rejects.
Adverse-action reasons and a full audit trail on every call — not a score with no rationale.
Decisions built on real-time API calls, verifiable and timestamped — never a model's best guess.
No licence, no implementation fee, no lock-in. You pay only for decisions actually rendered.
Compliance
Every decision ships with the evidence a regulator, a board, or a declined customer would ask for. The compliance work you'd otherwise resource is already in the product.
Why us
Former CEO of five listed banks, including Bank of Queensland. Has sat on both sides of credit and M&A.
Built the loan-origination and core-banking platform at a fast-growing FCA-authorised UK SME lender.
25+ years leading data and AI transformation across Europe, the US and Southeast Asia.
Four-time Australian Banking & Finance CRO of the Year; former Group Chief Risk Officer, Bank of Queensland.
The offer
We score a sample of your recent decisions — or run in parallel for a few weeks. You see decision time, approval uplift, and agreement rate against your current process, measured on your own book. No integration commitment, no cost.