Client: Neontri’s long-standing banking partner
Industry: Banking
Scope: Real-time transaction enrichment and integration with the bank’s systems
Key technologies: Snowdrop Solutions, Python 3, FastAPI, C++11, Apache Cassandra (DataStax), Apache Kafka, REST APIs, OpenShift, Terraform
Timeline: Approximately 3 months to implement the data management system
Challenge
The bank was receiving chargeback reports for legitimate purchases that customers couldn’t identify from their transaction descriptions. These mistaken reports added avoidable handling costs and customer service demand, while their volume put pressure on agreements with payment providers.
To address the problem, the project needed to:
- Identify why customers were filing erroneous reports.
- Reduce those reports within payment provider limits while maintaining customer service quality.
- Enrich large volumes of retail transactions in real time.
- Provide accurate merchant information so customers could recognize purchases.
- Meet data caching and storage requirements that varied by market and transaction type.
Are unrecognized payments driving disputes at your bank?
Solution
Neontri implemented banking transaction enrichment using Snowdrop Solutions, selected to meet the project’s requirements for data quality, cost optimization, and real-time processing.
Merchant names, logos, locations, and contact details added context to payment records, enabling customers to recognize purchases before raising a dispute. The approach was designed to reduce erroneous chargeback reports and improve the banking experience.
Technology
The technology choices supported high-performance data processing and flexible API integration:
- C++11: Processed and stored transaction data, with a focus on performance.
- Python 3 and FastAPI: Supported API integrations that could be adapted as requirements changed.
The wider stack included Apache Cassandra, Apache Kafka, REST APIs, OpenShift, and Terraform.
Delivery approach
Neontri worked with its long-standing banking partner in two-week sprints. Three developers and a project manager delivered the implementation, while the bank handled acceptance testing. The teams adjusted requirements based on performance testing.
A key part of the work was tailoring how transaction data was cached and stored to meet different legal requirements across markets and purchase types. Implementing the data management system took approximately three months.
Results
The project delivered benefits for both the bank and its customers:
- Reduced erroneous chargebacks: Real-time merchant details made purchases recognizable, reducing mistaken disputes and related customer service requests.
- Lower handling costs: Managing reports through payment providers and support teams became less costly.
- Stronger digital offering: Transaction enrichment improved the customer experience and strengthened the bank’s competitive position.
Reduce mistaken disputes with enriched transaction data
Add merchant names, locations, and purchase context to transaction records so customers can identify payments faster and your teams can lower unnecessary handling costs.
Our process for strategic partnership
-
01
Discovery
Business priorities, transaction data gaps, and regulatory requirements are assessed to define the project scope.
-
02
Design
The architecture is designed for reliable integrations, real-time processing, and compliant data handling.
-
03
Development
The solution is built around data quality, performance, and flexibility as requirements evolve.
-
04
Testing
Processing accuracy, integration reliability, and performance are verified against agreed requirements.
-
05
Deployment
The solution is introduced into the banking environment with a focus on operational continuity.
-
06
Optimization
Testing insights and client feedback guide refinements to data quality and system performance.
Paulina Twarogal
Marcin Dobosz
Banking Success with GenAI
