Automating the Credit Assessment Process
Country
Poland
Industry
fintech
A comprehensive redesign of the credit analyst working environment at a Polish commercial bank, replacing a fragmented multi-system workflow with a unified Client Card in Microsoft Dynamics 365 and a Generative AI component for automated company verification and risk recommendations.
Services
Technologies
Project in numbers
Reduction in actions
Fewer app switches
Faster analysis
Saved per month
Background
A Polish bank serving retail customers and the SME segment. Their credit analysts evaluate an average of 590 loan applications per month.
The project aimed to redesign the analyst working environment for the Micro credit process, covering discovery workshops, UX design, a Client Card in Microsoft Dynamics 365, and a Generative AI component.
The Challenges
Analysts worked across multiple parallel tools with no central analysis point — up to 96 application switches per case. Data was scattered across BIK, DEF, CEIDG, SUDOP, MEF, and KRS/REGON registries.
All verifications were manual. Average analysis time: 90 minutes per case, totalling 885 hours of analyst work monthly. No AI tools existed to support decision-making or generate analytical notes.
The Solution
A three-layer solution combining new data architecture, a redesigned analyst interface, and Generative AI.
The Client Card in MS Dynamics 365 aggregates all data in a single view: persistent header bar, automated rule verification with status indicators, historical data comparison, and synthesised analytical notes.
The AIConsole component automates registry checks, detects negative media, verifies affiliations, and generates analytical notes covering the full research cycle.
Business Impact
Freeing 698 analyst hours monthly creates capacity to grow without increasing headcount. Analysis time dropped from 90 to 19 minutes per case.
The modular architecture enables gradual expansion to the SME segment and other teams. AIConsole integration with the bank's internal assistant opens comprehensive conversational support for analysts at every stage.
Check other case studies
ETS Language Muse
Language Muse is an NLP-based educational platform for teachers working with U.S. students who face the challenge of learning English as their second language.
TimeION
TimeION simplifies the time off request policy and makes the experience pleasurable for both employees and managers.
GoSeqIt
10Clouds provided a platform that allows conducting DNA analysis through the internet, from any place on Earth. It computes and stores all data in the cloud and generates legible reports.
Zapp RideShare
Zapp RideShare is an electric scooter rental service operated with a mobile app. Our team is responsible for the complete product makeover.
