I am a Java backend engineer with 3+ years of experience building and maintaining systems in the forex and cross-border payments domain at First National Bank (FNB).

I work on backend services supporting real-time cross-border payments and FNB's migration from SWIFT MT to ISO 20022 MX messaging, where correctness, schema strictness and message integrity are non-negotiable. My day-to-day stack is Spring Boot and Java EE (with Apache Wicket still doing real work in production), and I design APIs that hold up in distributed, high-throughput environments.

I have practical experience with transaction boundaries, message queue reliability, and system behaviour under load, and I apply this when designing and improving backend systems.

• Built and maintained Spring Boot microservices for FNB's real-time cross-border payments platform (TCIB), now integrated with three regional banking partners across the SADC region
• Delivered core services for FNB's bank-wide migration from SWIFT MT to ISO 20022 MX messaging, including Spring Boot transformation microservices with XSD validation enforced at the inbound boundary
• Built the Spring Boot API connecting FNB's Illicit Flow Fraud Model into the live payments pipeline, which I also support in production
• Carry an on-call production support rotation and mentor two junior developers through fortnightly code reviews
• Work daily with REST and SOAP APIs, XML/XSD and JAXB, given how much of payments messaging still runs on strict schemas
• Containerised services using Docker for consistent deployment

Backend

• Java 8, 11, 17
• Spring Boot, Java EE, Apache Wicket
• REST and SOAP APIs, Microservices

Data and Persistence

• MySQL, Oracle SQL
• JPA/Hibernate, HQL

Messaging and Integration

• Apache ActiveMQ
• XML/XSD, JAXB
• OAuth 2.0, SSL/TLS

Infrastructure

• Docker, Linux
• Git, Maven

Testing

• JUnit, Mockito

JPA-aware fake data seeding library for Spring Boot, published to Maven Central.

Key features:

• Generates realistic seed data directly from JPA entity metadata
• Spring Boot autoconfiguration for drop-in setup
• Published under groupId io.github.thembatman0, MIT licensed

Maven Central: io.github.thembatman0:synthforge-spring Repository: ThembaTman0/synthforge

Microservices-based banking backend demonstrating service discovery, centralised configuration, and containerised deployment.

Key features:

• Service discovery using Eureka
• Centralised configuration
• Docker-based deployment
• Microservice-based architecture

Repository: ThembaTman0/Bankwave-V-2.0

Satellite Image Classification Using HOG and DAISY Feature Descriptors, co-authored with Ritesh Ajoodha. Published at the 2022 International Conference on Electrical, Computer, Communications and Mechatronics Engineering (ICECCME), Maldives, November 2022.

A classical computer vision pipeline, not deep learning: DAISY and HOG feature descriptors combined through Bag of Features encoding, classified with an SVM across 21 scene categories on the UC Merced Land Use dataset.

View on IEEE Xplore

If your synthforge or satellite-classification repo has its own README with a results table, that table still needs the same fix: change the "This work (SVM + RBF) 2022" row from 81.42% / 68.02% to 81.76% (hybrid) / 67.14% (KNN), matching the published abstract. That table isn't part of this profile README, so it wasn't touched here.

thembangobeni.com

_{Public repositories do not reflect private enterprise work.}

Email: thembatman0@gmail.com LinkedIn: themba-ngobeni-6a163b164