Leading Jewelry Retailer Transforms Its IBM Sterling OMS Support with AI Engineer

A leading North American specialty jewelry retailer runs one of the industry’s most complex omnichannel fulfillment networks—hundreds of stores, multiple distribution centers, and fulfillment paths spanning vendor drop-ship, ship-from-store, and same-day delivery. As order volume grew, the support operation behind it hadn’t kept pace: engineers were navigating four disconnected systems just to resolve a single incident, driving up resolution times and creating heavy dependency on a handful of senior experts

Nextuple built the answer: the AI Support Engineer, a conversational AI assistant that gives engineers one interface instead of four. It automatically routes every query to the right system—OMS data, database diagnostics, or operational runbooks—so engineers get answers, not a scavenger hunt.

The impact is immediate and compounding. Phase 1 alone cuts incident resolution time by ~30% and delivers a 10% to 15% cost reduction—funding the next phase from day one. By Phase 2, automation coverage reaches ~40% of total support activity, with 24/7 monitoring and no shift constraints. The result: a support team that spends less time hunting for answers and more time driving the business forward.

The retailer’s Sterling OMS production support operation had outgrown its tooling. Order exception volumes were growing alongside the business, but the support workflow had not evolved—leaving engineers dependent on a fragmented set of systems with no unified interface.

Support engineers simultaneously navigated Sterling OMS consoles, IBM DB query tools, shared documentation repositories, and ticketing systems to resolve a single incident. This context-switching drove up mean time to resolution (MTTR) and created structural dependency on senior engineer institutional knowledge—making onboarding slow and escalation rates high.

An analysis of support effort revealed the majority of engineer time was consumed by activities directly addressable through automation:

Nextuple delivered the AI Support Engineer—a purpose-built, multi-turn AI assistant integrated directly into the Sterling OMS production support workflow. Built on Azure OpenAI GPT-4o and deployed via a React-based web interface, it gives engineers a single conversational interface that autonomously routes queries to the right tool: Sterling OMS REST API, IBM DB Query Servlet, or a semantically indexed SOP knowledge base.

Roughly 30% to 40% of total OMS support activities are strong candidates for automation. The AI Support Engineer targets this directly—shifting teams from reactive issue resolution to proactive, strategic value creation. Phase 1 goes live in under four weeks, establishing monitoring and reporting automation before more complex chatbot and order-action capabilities follow.

The AI Support Engineer delivers measurable improvements across cost, speed, and coverage from the first phase of deployment. Figures are based on observed outcomes from live OMS support engagements; actual results will vary based on team size, current FTE mix, and scope of automation.