Building a Modern AI-Ready Backend: Web MVC, JNI, WebFlux, and Python RPC

I designed and implemented a modular backend architecture that supports traditional Spring MVC services, native AI acceleration through JNI, reactive APIs using WebFlux, and remote Python-based inference. Through this project, I explored multiple approaches to bridging classical Java backend engineering with contemporary AI workloads.


Demo Video

Here is the complete demonstration integrating all four components.


1. Java Web MVC Core Library

The module focuses on building a clean and extensible Java MVC backend library.


2. JNI Bridge for AI Integration

To connect Java with native AI modules, I developed a Java Native Interface (JNI) layer.


3. Reactive API with Spring WebFlux

Next, I added a non-blocking reactive API layer using Spring WebFlux.


4. Python RPC Model Service

Finally, I created a Python RPC inference server.


Java + AI Backend Integration

Spring MVC & WebFlux

Python Inference & RPC

AI Infrastructure & System Design

Retrieval, Evaluation & Governance

External References