An 8-person platform team and a 22-person automation COE support a 14 million customer bank. We keep it running by betting on orchestration, resilient automation, and case-based workflows where linear diagrams break down.
We have been serious about RPA since 2016 and serious about orchestration since 2024. Our data warehouse has been serious about us since 2019. Here is what that looks like today.
Four flagship automations demonstrated live at UiPath DevCon 2026. All in production. All running on UiPath.
Our platform team rebuilt the bank's automation stack in 2023 around UiPath. Different teams use different UiPath products as front doors into the same orchestration engine. The fourth layer is newer and covers the humans who still need to be in the loop, just with better tools.
The backbone. Business analysts model in a BPMN-style designer inside UiPath. Developers author the same Orchestrator runtime using UiPath Maestro, a flow-based, code-near experience. Same engine. Same deployment pipeline. Same observability.
How we chose two front doorsFor the workflows that cannot be drawn as diagrams. Commercial lending, claims, internal investigations. UiPath Case Management gives us stages that hold structure and tasks that adapt on the fly, all on the same runtime as the orchestration layer above.
Commercial lending case studyOur 420 bots are authored in UiPath Studio. Integrations prefer APIs and fall back to UI automation when upstream systems drift. Our UiPath Coding Agent helps the COE author, debug, and maintain bots at scale. The agent drafts. The engineer ships.
Why we fall back to UINot every task runs in a pipeline. Our QA engineers, relationship bankers, and COE leads use UiPath Agentic Desktop to handle the work that lives between systems: reviewing test results, reconciling data across applications, and routing actions without leaving their current context. It reads what is on screen, acts across tools, and reports back.
See it in our QA workflowPrinted, laminated, signed by everyone. In every Accrual Bank platform team room. We update them about once every two years when someone has a better idea. We added one in April 2026.
One orchestration runtime. Different authoring experiences for different audiences. No second product for developers. No second product for analysts.
Linear processes get BPMN. Everything else gets case management. Forcing chaos into a diagram is how you get a 47-step diagram that nobody maintains.
APIs are faster when they exist and stable. UI automation is the safety net when vendors deprecate without warning. Every integration ships with both paths.
Our code-generating agent for bot authoring is a pair programmer, not a magic button. Our engineers review every change. Our agent ships faster than either alone.
Every bot ships with contract tests for its APIs and regression tests for its UI targets. The 420th test is as important as the 1st. We test Delegate-class changes end-to-end.
BPMN processes, flows, and bots all ship through the same CI. Same gates. Same rollback strategy. Same observability surface.
When work is inherently non-linear, give it stages and tasks. Commercial lending is a case. Claims is a case. Internal investigations are a case.
If only one person can ship a change safely, that change should not ship. Our 8 platform engineers are interchangeable, on purpose.
Not the other way around. When a business team needs something that is not on the roadmap and it is reasonable, we move the roadmap.
Reliability is the brief. Prestige is a nice-to-have. Our incidents are rare and our pages are boring. Both are on purpose. The moment a platform starts optimizing for interesting, it stops being trustworthy.
When something breaks, the coding agent investigates before a human does. It pulls the instance trace, walks the flow node by node, identifies the root cause, and proposes a fix. The engineer reviews and merges. We do not skip the agent step because we are in a hurry. Especially when we are in a hurry.
Principle 10.5, unofficial: if an RPA bot is doing data entry and data validation and data extraction on the same screen, that is not a bot, that is a full-time employee in disguise.
It did not happen overnight and we did not plan it this way. We just kept saying yes to the next thing and yes to writing down why.
Posts from our platform engineers and COE leads. No sponsored content. No thought leadership. Just working notes from people building this stuff.
How an agentic desktop handles the full failure loop across Orchestrator, Test Manager, and Jira without a human switching tabs.
How our bots adapt when upstream dependencies break, using a hybrid pattern that prefers APIs and falls back to UI automation.
Why we gave our developers a flow-based designer on the same orchestration runtime we use for BPMN. And how it changed our velocity.
What the board asked us about AI, and what we told them. The strategy that came out of a weekend of honest writing.
Automation infrastructure
Every bot, every flow, every case, every agent at Accrual Bank runs on UiPath. Maestro for flow orchestration. Case Management for adaptive workflows. Studio for bot authoring. Agentic Desktop for humans between systems. One platform, one runtime, one bank.
If you like working on small teams that ship things used by millions of people the next morning, get in touch.