Python-Powered Crypto Banking Platform
- The evolution of a cutting-edge banking platform was handled by our Python developers, showcasing their prowess in creating solutions for QIWI, non-bank entities, crypto projects, and media players. The platform's core features included virtual Visa/Mastercard cards, and our team ambitiously planned over 70 BINs across diverse segments and geographies. Future enhancements will encompass proprietary crypto-processing capabilities and a robust core banking platform tailored for European and Southeast Asian markets., including exchange rates, core data ingestion mediation, FKG mediation, and background processing. The system was built to handle millions of transactions daily, with a focus on real-time processing and analytics.
Client
The client is a leading financial services provider, known for its extensive network of banking institutions and comprehensive financial products. They faced challenges with their existing transaction monitoring system, which struggled with scalability and integration issues. The customer required a modern, scalable solution to improve transaction processing speed, accuracy, and reliability, while also enhancing data security and compliance with financial regulations.
Detailed information about the client cannot be disclosed under the provisions of the NDA.
Solution
The primary application, initially a monolith, encompassed crucial functionalities like authentication, user management, card services, and transaction handling. A key microservice, developed using FastAPI and powered by PostgreSQL, facilitated seamless communication with other services via HTTP requests.
MassPay:
Handled mass payments to cards, SWIFT, and SEPA, including internal data processing, beneficiary validation, and interactions with providers like Fenige and NIUM. It was implemented using Go and Python with gRPC for communication and Temporal for workflow.
Financial Module:
Managed ledger interactions, account-level transaction processing, and report generation, utilizing Python, gRPC, and Postgres.
Results
75%
decrease of support requests
40%
decrease of resolution time
20%
improvement of system performance
35%
decrease of Fraud cases
25%
improvement of system security
Documentation:
Documentation creation time was reduced to 5-10 minutes; support requests decreased by 75%
Sentry and Grafana:
Error detection and resolution time decreased by 40%; average bug resolution time dropped from 4-5 hours to 2-3
Webhook Microservice:
Webhook processing time decreased from 5 seconds to 1; errors and failures reduced by 30%; system performance was improved by 20%.
Authorization Control:
Fraud cases decreased by 35%; system security improved by 25%.
Mass Payout Service:
Request processing time dropped from over 3 seconds to under 500 ms; the number of errors was reduced by more than half
Team
1
Team Lead
1
PM
4
Backend Engineers
2
Frontend Engineers
1
UX/UI designer
1
Data Engineers
1
Data Scientist
1
QA
1
BA
Technical Stack:
Backend
Frontend
Platforms
Conclusion
The successful evolution of our project into a crypto-friendly banking platform featuring virtual Visa/Mastercard cards was made possible by the efforts of our Python team at Pynest. The extensive future plans, including 70+ BINs for any segment and geography, proprietary crypto-processing, and a core banking platform for European and Southeast Asian markets, were all supported by our skilled developers.
