Snapshot

At Acko, I worked on workflow-driven backend systems for health insurance operations. My work spanned two tightly connected layers.

First, I built a focused automation system for pre-policy medical checkups, a critical flow that was slowing down policy issuance. This system reduced turnaround time from 17 days to 3 days.

Second, while solving that problem, I identified a broader architectural gap: several insurance processes were long-running, stateful, exception-heavy, and difficult to manage cleanly through scattered service logic. I proposed a workflow-first architecture using Camunda, BPMN, and DMN, backed by microservices. That approach proved useful beyond a single use case and influenced broader insurance workflow design.

Business Context

Health insurance operations are rarely simple request-response flows. Real journeys involve:

• multiple dependent steps
• business rules and branching logic
• external and internal actors
• asynchronous progress over time
• retries, exceptions, and manual interventions
• operational visibility requirements
• auditability and state clarity

A strong example was the pre-policy medical checkup flow. In certain cases, the user could not move straight to policy issuance. The application had to pass through underwriting-related checks, medical coordination, status updates, and downstream completion steps before the policy journey could continue.

The business problem was not only delay. It was structural fragility.

The Problem

Before automation, these flows were hard to manage because the process itself was not treated as a first-class system concern.

Typical issues included: