Most enterprise AI stories still sound the same: a new copilot, a promising pilot, and a lot of language about transformation. What is harder to find is a case where AI is attached to a painful operating metric that a business already cares about. That is why the latest Xpress Boats example is worth attention.
In April 2026, AWS and Infor highlighted Xpress Boats as an early manufacturing customer using Infor Velocity Suite, Process Mining, automation, and generative AI-driven document handling to improve real workflows across procure-to-pay, order-to-cash, and demand-to-build. According to the vendor-published case, Xpress Boats improved process issue diagnosis speed by 98%, cut returns processing time by 95%, and reduced expedited shipping costs by 50%.
Those are not vanity metrics. In a manufacturing business, expedited shipping is usually the tax you pay for weak planning, poor visibility, and late decisions. If AI helps reduce that cost materially, it is not just making work feel faster. It is changing the economics of operations.
Why This Case Matters
Xpress Boats manufactures aluminum fishing boats, pontoon boats, and trailers out of Hot Springs, Arkansas. Its competitive edge depends on getting products out on time, which means the company lives or dies on operational coordination. That matters because AI is often weakest in messy, cross-functional environments where no single team owns the entire bottleneck.
This case is interesting precisely because the company did not start with a broad ambition like “use AI across the enterprise.” It appears to have started with the workflows where missed signals become expensive. Infor says the system surfaced bottlenecks in less than a week, then paired those insights with automations for returns processing, vendor pricing, and document-heavy work. That sequence is important: first diagnose the friction, then automate the part that can be standardized, then optimize around the result.
That is much closer to how successful AI adoption usually works in the real world. PwC’s 2026 AI Performance Study found that the companies capturing the most value from AI are more likely to redesign workflows around it, not simply bolt AI onto existing habits. Xpress Boats looks like a concrete example of that pattern in manufacturing.
What Xpress Boats Actually Did
Based on the recent AWS and Infor material, the operating model looks fairly clear. Xpress Boats used Infor Process Mining to analyze workflow activity across three high-friction areas:
- Procure to Pay, where supplier timing and purchasing errors can cascade into shortages or rushed orders.
- Order to Cash, where delays and exceptions can slow fulfillment and customer communication.
- Demand to Build, where planning errors can disrupt production schedules and force reactive logistics decisions.
Infor says those insights were combined with automation for returns processing and vendor pricing, plus generative AI for intelligent document handling. The new agent layer is also being tested across purchase orders, customer orders, general ledger, accounts payable, and accounts receivable. In plain English, AI is not being used here as a standalone assistant. It is being embedded into the connective tissue of operations where slow diagnosis and repetitive exceptions create cost.
That distinction is easy to miss, but it is central. Many executives still think of enterprise AI as chat interfaces for knowledge workers. This case points in a different direction. The larger opportunity may be operational AI tied to workflow telemetry, where the system can spot process variants, recommend fixes, and automate pieces of the response.
Why The Metrics Are More Useful Than Typical AI Hype
The strongest number here is probably the 50% reduction in expedited shipping costs. Manufacturing leaders understand that expedited freight is rarely the root problem. It is the downstream symptom of planning failures, process delays, and last-minute recovery actions. If that cost is cut in half, something upstream likely improved in a meaningful way.
The 98% faster diagnosis figure matters for a related reason. AI creates disproportionate value when it shortens the time between a problem emerging and the organization understanding what is actually wrong. Faster root-cause detection means fewer hours wasted in escalation chains, fewer emergency workarounds, and fewer mistakes carried downstream into production and delivery.
The 95% faster returns processing claim also fits the same pattern. Returns and exception handling are exactly the kinds of ugly workflows that absorb employee time without creating strategic value. If AI can standardize those tasks, the gain is not just labor saved. It is a cleaner operating system for the business.
What Leaders Should Learn From It
The first lesson is that the best AI use cases usually hide inside expensive operational frictions. Xpress Boats did not need a grand reinvention story to create value. It needed faster visibility, better process compliance, and fewer reactive costs.
The second lesson is that AI works better when paired with system context. This case appears to depend on workflow data already living inside Infor’s manufacturing and ERP environment. That matters because AI without context usually produces weak recommendations, while AI grounded in real process data can identify specific failure patterns and automate specific fixes.
The third lesson is that narrow wins build the foundation for broader agent adoption. Infor says Xpress Boats is now testing agents across finance and order workflows. That sequence makes sense. Prove measurable value on one painful metric first, then expand into adjacent workflows once trust exists.
Successful AI adoption is usually less about intelligence in the abstract and more about removing delay, ambiguity, and rework from a business system.
The Caveats
The caveat is straightforward: this is still a vendor-published case study. Infor and AWS are both telling the story, and the reported metrics are not independently audited. That means the numbers should be treated as strong directional evidence, not final proof. It is also not fully clear how much of the gain comes from process mining, automation, generative AI, or the broader operational redesign around those tools.
Still, this case is more useful than most AI marketing because it names the workflows, identifies the operational problem, and reports metrics that matter to a real business. That is enough to make it strategically interesting even if the exact attribution remains fuzzy.
The Business Takeaway
Xpress Boats offers one of the clearer recent business cases for AI adoption because the target was not “productivity” in the abstract. The target was operational drag. By using AI to diagnose process issues faster, automate repetitive exceptions, and reduce emergency logistics costs, the company appears to have improved the part of the business where inefficiency is easiest to feel and hardest to hide.
If you run a manufacturing or distribution business, the practical takeaway is simple. Start with a workflow where failure creates a visible cost: expedited freight, return handling, order exceptions, planning delays, or invoice rework. Instrument the process, find the bottlenecks, and apply AI where it can reduce the time between signal and action. That is how AI stops being a pilot and starts becoming an operating lever.