9-day loan decision at a regional bank — agentic origination for small business
A regional bank was losing small business loans to fintechs because their decision cycle ran 21 days. pronix.ai deployed an agentic origination workflow with fair-lending guardrails and underwriter-in-the-loop — decisions in 9 days, 2.6x underwriter throughput and zero fair-lending findings.
- Client
- Regional US bank, $48B in assets
- Industry
- Financial Services
- Platform
- Salesforce Financial Services Cloud · AWS Bedrock · nCino
*Representative outcome; results vary by client, scope and platform configuration.
The challenge
The bank's small business origination was a 21-day relay across relationship managers, credit analysts, underwriters and documentation — losing 34% of qualified applicants to faster fintech competitors and unable to explain declines to fair-lending examiners.
Our approach
One agent per relay hand-off
Replaced hand-offs with bounded agents for intake, document extraction, spread analysis and covenant checks — humans owned decisions, agents owned prep.
Fair-lending guardrails
Every decision surface excluded protected-class proxies and produced a plain-language reason code aligned with ECOA and Reg B.
Underwriter cockpit
Underwriters saw the agent's proposed decision, its evidence chain and any exceptions — approve, modify or reject in one screen.
Portfolio-level backtest
Ran the workflow against 3 years of historical loans before launch — confirmed no adverse-impact drift versus prior process.
Continuous adverse-action monitoring
Weekly fair-lending report by product, geography and demographic proxy — surfaced to CRO before it surfaced to a regulator.
“We can now compete with the fintechs on speed without giving up the credit discipline that makes us a bank.”
Illustrative case study. Scenarios, metrics, quotes and client details are representative composites based on Pronix engagements and industry benchmarks unless a named client is shown with written consent. Outcomes vary by client, scope, data quality and platform configuration. Nothing on this page is a guarantee, warranty or professional advice. See our Terms of Use for the full disclaimer.
AWS, Salesforce, Microsoft, NICE, Genesys, Kore.ai, Google, ServiceNow, Amazon Connect and logos referenced on this site are trademarks of their respective owners. References are for descriptive purposes only and do not imply endorsement, sponsorship or partnership beyond stated partner relationships. See our Disclosures.
Talk to the team that shipped it.
Book a working session with a pronix.ai lead on this practice — we'll map the approach to your platform, industry and constraints.