AG.ALI GHNAINIA
Let's talk

CASE STUDY 01 / REAL-TIME SYSTEMS / MOBILITY

SAEIV — Real-Time Transport Supervision Platform

A production platform for supervising autonomous shuttle and transport operations in real time.

All projects
Spring Boot microservices
8+
records in source tables
22M+
new rows per day
50K+
connectors orchestrated
~42

ORCHESTRATION PATH

~42connectors orchestrated

  1. 01CSV ordering rules
  2. 02Connector JSON
  3. 03Kafka Connect REST API
  4. 04RUNNING status and lag checks
  5. 05Recovery around failed tasks
01

Problem

The platform needed to surface live transport information while moving high-volume changes from operational data sources into downstream services. Production startup also required careful sequencing of Kafka Connect and Debezium connectors.

02

Architecture

Spring Boot services communicate through REST APIs and Kafka. Spring Cloud Gateway, Eureka and OpenFeign support service access and discovery. Debezium and Kafka Connect capture database changes; Angular and WebSocket deliver real-time views to operators.

  • Secured REST endpoints with JWT
  • PostgreSQL source data and Kafka event streams
  • Docker Compose, GitHub Actions and Nginx for deployment
03

What I Built

I designed and developed more than eight Spring Boot microservices and built a standalone Java 17 / Spring Boot orchestrator for connector startup after deployment. It reads ordering rules from CSV and connector configurations from JSON, creates connectors through the Kafka Connect REST API, waits for RUNNING status, checks Kafka activity and consumer lag, then advances when conditions are safe.

04

Scale

The source tables contained more than 22 million records and grew by over 50,000 rows per day. Around 42 Kafka Connect / Debezium connectors required controlled creation and startup.

05

Technical Challenges

Connector tasks can fail or start out of order. The orchestrator tracks task state and supports recovery around failures, keeping the startup sequence observable and controlled. I also contributed to deployment follow-up and production issue investigation.

06

Result / Engineering Value

The startup process became a repeatable, rule-driven workflow with explicit health checks before each next step. I supported technical coordination, priorities, code reviews and developer support around the platform.

07

Tech Stack

  • Java 17
  • Spring Boot
  • Spring Cloud Gateway
  • Eureka
  • OpenFeign
  • REST APIs
  • Kafka
  • Debezium
  • Kafka Connect
  • CDC
  • Angular
  • WebSocket
  • JWT
  • PostgreSQL
  • Docker Compose
  • GitHub Actions
  • Nginx

NEXT PROJECT

STATLINK

A reporting platform that connects operational data, interactive dashboards and scheduled outputs.