All work
Zions Bancorporation · Banking · 2024 — 2025
AI-Assisted Transaction Monitoring
Event-driven banking services that score transactions with ML models for anomaly detection and risk insight — without slowing down payments.
Client code · private under NDA
Architecture at a glance
01
Banking app
React · customer + ops
02
Payments service
Spring Boot · JWT
03
Kafka
txn.created
04
ML scoring
Anomaly + risk
05
Ops dashboard
Insights & alerts
The challenge
Transactions had to be processed instantly and securely, while still being checked by ML models for anomalies and risk.
Tech stack
Java 17
Spring Boot
Spring Security
Apache Kafka
Python
React
PostgreSQL
Oracle
AWS
JUnit
Mockito
What I did
- 01Built Java 17 Spring Boot services for accounts, payments and transaction processing.
- 02Streamed transactions over Kafka to AI/ML services asynchronously, keeping the payment path fast.
- 03Validated and transformed transaction data for predictive models; surfaced insights in React operational dashboards.
- 04Locked down sensitive APIs with Spring Security, OAuth 2.0, JWT and RBAC; tuned PostgreSQL/Oracle for high volume.