A global energy major's Endur release pipeline was carrying too many hands. Every move from development through to production required manual intervention: environment setup, deployment steps, configuration, test process orchestration, scope decisions. Each release became a small project rather than a routine event.
Underneath the surface, test environments were drifting from production. Configurations diverged. Market data fell out of sync. The result was a release window that kept stretching, a cost-per-defect that kept climbing, and a team spending more time managing the pipeline than testing the platform.
The solution
The work focused on optimising side-by-side test automation across the Dev-to-Prod pipeline. Triangle sat in the test layer, doing the comparison-heavy lifting that side-by-side testing depends on: pre versus post environment comparison across trades, prices, simulations, reports and downstream outputs. Around it, deployment, infrastructure and test process steps were automated, removing the manual touchpoints between each stage.
More frequent production snapshots brought test environments closer to live behaviour, so the comparisons ran against data that actually reflected the platform's real state. Test scope was pulled forward into release planning, with the team deciding what to test at which stage rather than retrofitting after the fact.
The results
50% reduction in human touchpoints from Dev to Prod within 18 months, achieved through automation.
Manual interventions and handovers were removed across deployment, infrastructure and test process steps, freeing engineering time for higher-value work.
Closer alignment between test and production environments, with more frequent production snapshots reducing the noise from environment drift.
Test scope decisions became part of release planning, replacing reactive choices with deliberate ones.
Defect discovery became more targeted, with less extraneous noise to triage around each release.
Why Trinitatum?
Side-by-side is our specialism. Around 80% of Triangle's out-of-the-box content is oriented around comparison, which is what makes it practical to cover the full surface of a live Endur platform without writing every test by hand. The release pipeline doesn't need a tool that does a bit of everything. It needs one built for comparison work at scale. That's what Triangle is.
