CASE 02·Textiles · Functional fabrics

TEXRAY × Intellicon Solutions

More than 40 AI Agents in three months
MCP to break down data silos
and connect processes end to end

47 yearsIn business

3 monthsFrom process mapping to go-live

40+Built by the internal team

1 hour→minutesFabric inspection report handling

Before the rollout

TEXRAY (1467), founded 47 years ago, is a textile and apparel manufacturer with production sites on three continents. That global footprint brings a hidden difficulty: every country wants different customs codes and packaging rules. Staff had to work like detectives, comparing thousands of records back and forth across PDFs, Excel and ERP. Confirming a single shipping document could take half a day.

Repetitive administration of this kind eats hours and wears down creativity. TexRay understood that staying competitive internationally meant using technology to break down the data silos and bring workflows worldwide to one standard.

Goals

Eliminate the time spent comparing documents and entering data by hand
Build a bottom-up culture of AI use and grow an internal AI team
Cut the human errors caused by complicated cross-border document rules
Use MCP to connect AI Agents deeply with existing systems
Turn tacit knowledge such as fabric inspection judgment and customs declaration logic into digital assets

What Intellicon Solutions did

TexRay took a bottom-up approach, picking seed members from sales, shipping and quality control. These were people who already thought in structured terms, and they picked up the development skills quickly.

Then came MCP (Model Context Protocol), which took the AI beyond answering questions to actually reading from and writing to the company’s internal databases and systems:

01Mapping every process: more than 40 tasks suited to working with AI identified within 3 months
02MCP system integration: give the Agents read and write access to databases, connecting systems that had each run on their own
03Seed team training: staff who know the textile business design the Agents themselves, which is what makes them usable
04Regular review meetings: milestones and issue tracking keep performance improving

Use cases: 40+ AI Agents at work

Quality assurance (fabric inspection report Agent): the AI reads handwritten or PDF inspection reports, calculates defect points using the company’s own formula, and writes the result into the quality management database and Excel sheets through MCP.
Shipping (shipping document Agent): reads orders, checks them against each country’s customs codes and packaging rules, and generates the correct Shipping Instruction and Invoice.
Sales (order tracking Agent): compiles the status of cross-border orders automatically, cutting the time spent confirming them by hand.

Results

Development speed: in just 3 months the internal team built more than 40 AI Agents on its own
A leap in efficiency: handling a fabric inspection report went from 1 hour to a few minutes
System integration: MCP lets the AI read and write across systems like a “digital colleague”
Cultural change: the Agents are designed by the people in sales, shipping and quality control, not built for them by IT
Errors to zero: far fewer paperwork errors from checking complicated customs rules by hand

What the management team says

“For technology to take hold, the people who use it have to take part in designing it. We decided from the start that an AI project could not belong to IT alone. When someone builds an AI Agent with their own hands that saves them hours of work, the sense of achievement and ownership that comes with it beats anything a top-down project can produce.”

— 林宗頤, President, TexRay

What the partner says

“Even after working with hundreds of companies, we were still struck by how efficiently TexRay rolled this out and by what it produced. Their people have strong ideas about how to apply AI Agents and the drive to follow through. It is no exaggeration to say that TexRay does not look like the garment and textile industry people picture. It looks more like a technology startup.”

— Claudia Chen, COO, Intellicon Solutions

What’s next

In 2026 TexRay plans to extend AI Agent coverage across the full value chain, from R&D and procurement through production to shipping.

After that comes a network of collaborating Agents: when the quality control Agent finds an anomaly, it triggers the supplier communication Agent to send a notice and the production scheduling Agent to adjust the plan.

What TexRay shows is this: a traditional industry that uses AI and MCP deeply enough reaches the same efficiency and flexibility as the technology sector.