A significant enhancement to ESG Global's Business Orchestration Layer Engine — introducing BPMN-driven processes with Activiti and resolving a critical performance bottleneck that unlocked 30% additional overnight message processing capacity.
ESG Global's Business Orchestration Layer Engine — the middleware through which millions of smart meter messages flow daily — was hitting a processing ceiling during overnight read windows. Energy clients were constrained by capacity limits that couldn't be solved by adding hardware.
Profiled and identified a structural inefficiency in how the Activiti BPMN engine managed process instance state under sustained load. Resolved at the architectural level rather than tuning around it. Introduced new BPMN-driven business processes alongside the performance fix, extending platform capability without increasing operational complexity.
30% increase in overnight message processing capacity within the same infrastructure footprint — directly expanding the number of smart meter customers energy suppliers could serve. New BPMN processes extended the platform's capability while keeping orchestration logic transparent and auditable.
"Samuel's technical expertise and professionalism at ESG Global elevated our entire development process. His strategic insights, deep knowledge of smart metering systems, and consistent delivery drove meaningful and measurable improvements across our platform."
ESG Global — Smart Metering & Energy PlatformESG Global sits at the heart of the UK’s smart metering rollout, providing the software infrastructure that facilitates data exchange between energy suppliers and the Data Communications Company (DCC). Their Business Orchestration Layer (BOL) Engine is the critical middleware through which that data flows — handling millions of meter read and command messages daily.
My engagement at ESG Global ran from 2020 to 2022, and centred on two areas: extending the BOL Engine with new business processes, and resolving a performance bottleneck that was constraining overnight meter read capacity for major energy clients.
The overnight meter read window is time-constrained — energy suppliers have a fixed window to collect read data from millions of smart meters, and any capacity ceiling directly limits how many customers can be served. When I joined the team, the engine was hitting a processing ceiling that couldn’t be addressed by simply adding hardware.
Through profiling and analysis, I identified the root cause — a structural inefficiency in how the Activiti BPMN process engine was managing process instance state under sustained load. Resolving this unlocked 30% additional message processing capacity within the same infrastructure footprint, directly benefiting the energy clients relying on the system for their overnight read cycles.
The BOL Engine was built on Java 11 and Spring, with Hibernate for persistence against MS SQL Server, RabbitMQ for messaging, and Apache NiFi for data transformation pipelines. Infrastructure was managed through Maven, Jenkins, and Nexus, with GitFlow branching in GitHub and all project documentation maintained in Confluence.
The performance work delivered immediate, measurable value for ESG’s energy clients — expanded overnight capacity without any infrastructure cost increase. The BPMN process additions extended the platform’s capability while keeping the codebase maintainable and the orchestration logic transparent and auditable.
ESG Global Activiti BPMN RabbitMQ Apache NiFi MS SQL Server Spring JUnit 5 Mockito