Case study: Real-time Monitoring Engine

Case Study

Visual Basic out. Java in. Up to 70% faster.

Challenges
Project Journey
Results

Reduced processing time from 32 to 10 hours

~70%

Monthly savings Reduction of infrastructure costs

€10.000

Reduced maintenance effort Consolidated 3 Kubernetes Pods into 1

3 auf 1

LEAN-CODERS optimized thecritical Core Processing Engineof an internationally leading manufacturer of gas engines to monitor thousands of power generators worldwide. Througharchitecture modernizationandcode consolidationwe were able to reduce the processing time from32 to under 10 hoursreduce.

Challenges

The client faced critical challenges in their Reliability Engineering.

Performance bottleneck

The existing Core Processing Engine for monitoring global power generator fleets took 32 hours for daily data processing – far too slow for modern requirements.
Legacy architecture

Outdated Visual Basic code with complex network handovers between isolated systems, hard to maintain and prone to errors.
Lack of transparency

Millions of machine messages daily, but no efficient method for validating state changes and algorithms.
Infrastructure overhead

Three separate Kubernetes Pods for processing, high ongoing costs.

Project Journey

Analysis & Simulation

We built a comprehensive simulation environment with JavaScript (Node.js) and complete unit tests. Millions of messages could be sent through old and new systems to precisely identify differences.

This investment paid off– we discovered that the originally desired Python solution would have been even slower than the legacy system.

Architecture redesign

The client was already using Java for parts of the processing. We strategically expanded this existing Java architecture and migrated the critical processing logic from Visual Basic.

The breakthrough:We eliminated the network handovers between the systems by consolidating everything into a single, optimized Java codebase.

Implementation & Validation

A LEAN-CODERS Senior Developer worked closely with two experts on the client side. Each optimization was tested against real production data.

The result:fully documented, maintainable Java code with comprehensive unit tests – readable for the entire development team.

Deployment & Maintenance

Seamless rollout without downtime.The new Core Processing Engine has been running stably in production since then and is continuously maintained and optimized by us.

Results

Long-term partnership:

The project is one of several successful projects in a multi-year collaboration. Continuous maintenance is ongoing to this day.

We deliver. They stay. Repeat.

Performance

70% faster processing: From 32 hours to under 10 hours per calculation cycle

Consolidated architecture: Reduced three separate Kubernetes Pods to a single one

Costs & Maintainability

~€10,000+ monthly savings through infrastructure optimization (estimated)

Drastic reduction in helpdesk calls thanks to more precise algorithms and better reliability

Fully documented in Confluence with comprehensive unit tests

Technical quality

Modern Java codebase instead of hard-to-maintain Visual Basic legacy code

Foundation for follow-up projects

The new architecture already enabled a second project for other teams in the company

Team-ready

Any Java developer can understand and extend the code