Order Management System Modernization

Order state was tracked across multiple SOP (sub-order processing) systems with no single aggregated view at the customer-order level. Customers, customer-service agents, and downstream analytics all worked from inconsistent state, with no shared definition of what "this order" actually looked like once it split across sub-orders.

In parallel, the client was bringing a major new marketplace partner (MIRAKL) online and needed end-to-end testing coverage across the partner integration before turning third-party order flow on in production. The OMS team had deep platform knowledge but limited senior bandwidth — a classic single-engineer-deep risk profile for a system this central to revenue.

A senior backend engineer embedded into the OMS team and took ownership of two parallel workstreams.

The first workstream redesigned order-status aggregation. We built a new aggregation layer in the OMS database that rolled SOP-level statuses up into a single customer-facing order state, and introduced new optimistic / progressive status definitions so that in-flight orders surfaced sensible state to customers and CS agents even before all sub-orders finalized.

The second workstream took technical ownership of the MIRAKL partner integration. We subscribed the OMS to upstream Kafka topics, processed message flows, and drove end-to-end testing through multiple defect-resolution cycles to drive the integration to release-readiness.

In parallel, we contributed to the client's broader AI research spike — evaluating where generative and predictive techniques could fit into OMS workflows — and participated in PI planning to help the team frame two-month delivery objectives.