Your underwriters decide. Zoral prepares the file.

No. Your credit policy sets which applications can be decided automatically and which must go to a person. Your lending authorities stay as they are. Underwriters can override a calculation or recommendation within their authority, with the reason recorded. A customer who asks for a person to review an automated decision can be routed to an underwriter. Your people judge the exceptions.

Each extracted value links back to the document, page and field it came from. Each calculation follows rules your credit team has approved. The underwriter sees the source next to the result and can correct it. A correction creates a new version, so both the original and the change are kept.

Zoral looks for signs of editing within each document and for inconsistencies between documents, the application and external data. A flag is a reason to look, not a verdict. It goes to a named person with the evidence, and your policy decides what happens next.

Machine-learning models help predict default and inform risk-based pricing and limits. Language models help read documents and draft summaries and the arguments for and against an exception. AI agents do this inside the screens your underwriters already use, configured to your policies and workflows. Each agent has its own instructions, guardrails, budget and approval steps. Agents act only through the permissions, workflows and approvals you set, and every prompt, tool call and answer is traced for review. You choose the language models, from enterprise model services to open-weight models hosted in your own environment. Your knowledge bases, conversation records and client data stay under your governance, within the environments and agreements you control. Your own AI agents can use Zoral fOS through a standard interface (MCP), under the same controls. Agents do not make credit decisions. Automated credit outcomes follow rules and models your bank has approved. AI for Zoral helps our experts and your teams design and test them. People approve.

Yes. Each override is kept as a new version beside the original figure, with the person’s name, the reason and the time. The file also keeps the documents, the policy and model versions in force and the reasons for the outcome. Quality reviewers, auditors and supervisors can see how the file was decided, and by whom. We have worked with clients regulated by the FCA in the UK, the CFPB in the US and the central banks of Malaysia and Thailand. The obligations stay yours. Zoral gives you the evidence.

Models are versioned, tested on past applications before release, and can run alongside the current model on a share of applications before replacing it. Performance and drift are monitored against what you expected. This supports your own model risk management. Your model owners and validators keep the final say.

It doesn’t have to. Zoral can work alongside an existing origination system, including widely used mortgage origination systems. It reads the loan file and writes results back. Lenders who also want origination can run both on Zoral. You decide which system does what.

Both. For small businesses, Zoral analyses open-banking and bank statement data, connects to common bookkeeping systems, checks guarantors and collateral, and raises the conditions your policy requires. For larger businesses, it reads annual reports including the notes, spreads the financials and drafts the credit memorandum. Relationship managers and approvers keep the decision.

Underwriters stop re-keying, recalculating and chasing conditions. Their time goes on the files that need judgement. Each underwriter can handle more files without lowering the bar. Losses are compared with plan, so you can check that decisions are consistent. We measure the effect on your own volumes before you extend it. You make the staffing and authority decisions.

Zoral runs in your chosen region on AWS or Azure, or in your own data centre, to suit data residency rules. Zoral Document Fabric, which reads your documents, runs in your own AWS or Azure account, or your private cloud, in your region and under your control. Client data stays under your governance, within the environments and agreements you control, including for the language models you connect. Zoral connects to bureaus, open-banking providers and your own systems through documented interfaces. It is designed for high availability. Our controls are independently audited under SOC 2, with details on the Zoral Trust Portal.

You choose. Zoral can run and support the platform for you, your own technology team can run it, or we share the work and hand over as your team grows. Your people run the banking operations, including every credit decision that needs a person.

We usually start with one product and the step where files wait longest, such as income analysis, document checks or SME credit papers. Once we have scoped it with your credit team, we set out a realistic plan. We then extend to further products in regular releases. Your data, rules and audit history remain yours and can be exported in standard formats.

Zoral is an independent specialist, founded in 2004, with over 100 projects and 350+ years of combined experience. We have built credit models and automated underwriting for lenders since long before the current interest in AI. We design affordability into consumer credit. We work alongside your advisers and integrators, have a long-standing relationship with KPMG, and offer references in confidence.

Atelier, Zoral’s consulting and implementation team, works with your teams and, where you choose, with your advisers and systems integrators. The same team includes credit specialists who can review your scorecards and decision rules with your risk team.

Credit risk and decisioning