A textile manufacturer serving a nationwide network of B2B customers was running its entire order operation on fax machines and phone calls. Orders arrived throughout the day, were transcribed manually into the company's ERP by a team of staff, and customers had no visibility whatsoever — no order confirmation, no production status, no delivery date, no invoice access — without picking up the phone.
The operational consequences ran deeper than customer frustration. Because demand arrived in fragments and was entered manually, management had no consolidated picture of what had actually been ordered at any point in time. Decisions about what to take into production — which products, which quantities, when to schedule runs — were being made on instinct and incomplete data. The result was a chronic cycle of overproduction that locked up warehouse space and manufacturing capacity, and underproduction that meant missed windows and lost revenue.
The Challenge: A Bridge, Not a Replacement
The ERP at the centre of the operation was the company's system of record and could not be replaced. It was deeply configured to the business's compliance and operational requirements, and modifying it directly carried significant risk to the data integrity the business depended on.
The challenge was not to build a new ERP. It was to build a cognitive bridge: a modern, purpose-designed layer that sat in front of the legacy system, gave customers a self-service interface they could actually use, and made the data inside the ERP visible and actionable for management in real time — without a single change to the underlying system.
The integration had to be bidirectional and instantaneous. A customer placing an order needed to create an ERP record immediately — not in a batch, not at end of day. And every ERP state change — a production update, a dispatch event, an invoice raised — needed to reach the customer's portal view without any human in the loop.
The Architecture: Sync · Stream · Scale
We applied our three-stage framework to this engagement. Each stage addressed a distinct layer of the problem.
01 — Sync: The Data Liquidity Layer
Before a single customer-facing screen was designed, we solved the data problem. We built a real-time bidirectional API integration layer between Ruby Connect and the legacy ERP — engineered so that data that had been locked inside the ERP became live, accessible, and actionable across the new system instantly. Static operational data became liquid. The ERP remained the single source of truth. Ruby Connect became the lens through which that truth was made visible.
02 — Stream: Autonomous Order Processing
We replaced the manual transcription desk with an autonomous order stream. Ruby Connect was built as a purpose-engineered web platform — designed for performance at scale and seamless integration with the legacy ERP — giving B2B customers a clean, task-oriented portal to place orders, check production status, view dispatch updates, and manage invoices without contacting staff.
Every customer action triggered an immediate ERP record. The portal handled MOQ logic automatically — alerting the production team when accumulated orders for a product reached the minimum order quantity threshold that made a run economically viable. The logic that previously required human judgement and incomplete data now ran continuously and accurately in the background. 100% of the manual order-entry process was eliminated.
03 — Scale: Intelligence-Driven Production
With clean, real-time demand data flowing for the first time, we built the management intelligence layer: a suite of MIS reports covering aggregate demand, MOQ tracking, delivery schedule visualisation, and inventory commitment analysis.
For the first time in the company's history, production decisions were driven by data fact rather than instinct. Management could see exactly what had been ordered, by whom, for when — and plan manufacturing output to match actual demand rather than estimated demand. Overproduction and underproduction both became measurable and addressable rather than accepted operational realities.
What Was Left Standing
Ruby Connect is now described internally as the heart and soul of operations at the company.
The fax machines are gone. The phone queue for order status is gone. The end-of-day transcription backlog is gone. B2B customers nationwide now self-serve their entire account relationship through the portal — placing orders, tracking production, reviewing invoices, and managing delivery windows without contacting a single staff member for routine queries.
Management makes production decisions based on real-time consolidated demand data. Overproduction and underproduction have both been materially reduced. Warehouse capacity is planned rather than discovered. The MOQ reports that were previously impossible to produce accurately are now generated automatically and reviewed as part of the standard production planning cycle.
The most telling measure of the system's success came from an unexpected direction: both customers and internal staff now find the ERP significantly harder to use than Ruby Connect. The portal built to sit in front of the legacy system has become the preferred interface for the people who run the business.
The Lesson for 2026
The Ruby Connect engagement proves a point that becomes more relevant as AI adoption accelerates: you do not need to destroy your legacy systems to build modern capability around them. By establishing a clean data integration layer first (Sync), automating the operational workflow that depends on it (Stream), and then building the intelligence layer on top of that clean foundation (Scale), you create a platform that compounds in value over time.
The legacy infrastructure at this manufacturer remains untouched. The intelligence built around it made it irrelevant as a constraint — and powerful as a foundation.
Sources
- En Interactive Technologies — Ruby Connect engagement documentation
- Client operational data — provided under confidentiality agreement