The Watch Tower Blog
AI for Supply Chain, Built With the Suppliers You Know
AI for supply chain works when you build it with your buyers and suppliers. A metal fabricator shows how, and what peers at your size are doing.
You want material to show up when the job needs it, at the cost you planned, and you want to know early when it won't. That is the goal behind every conversation about AI for supply chain, and it is a goal no business reaches alone. Your supply chain is your suppliers, your carriers and the buyers who work with them. So the question that holds most operators back is not about the technology. It is: are businesses like mine actually doing this, and is it working for them?
That is the third of the five barriers between AI and real results, Ecosystem: are people like me using it, and is it working for them? (All five are in how to start using AI.) Without peers, suppliers or examples nearby, AI feels like a risk taken alone.
Where businesses like yours actually are
The honest answer is that most of your peers are early too. The U.S. Census Bureau's Business Trends and Outlook Survey found that overall AI use by U.S. businesses hovered between 17% and 20% from December 2025 to May 2026, with use higher at large firms: in the same release, 37% of firms with at least 250 employees reported using AI, and fewer than one in five of the smallest firms did.
Two things follow. First, if you have not put AI to work in purchasing or operations yet, you are in the majority, not behind it. Second, there are few worked examples nearby, especially in industrial supply chains. That is exactly why building it with the people around you matters more than picking the right product.
A worked example: a fabricator and its five critical suppliers
Here is an illustration, not a customer story. Picture an 80-person structural and miscellaneous steel fabricator. It buys plate and shapes from two service centers, fasteners from one distributor, coatings from one supplier and galvanizing from an outside galvanizer. Its customers are general contractors with hard schedule dates.
The purchasing lead's week is spent on the phone. When a galvanizer's turnaround slips, she hears about it when the steel doesn't come back. When a service center is short on a size, she learns at order time. Every late delivery becomes a late job, and the project managers find out last.
The plant manager wants to "use AI for the supply chain". The danger is buying a tool, connecting it to the purchasing spreadsheet and expecting it to predict delays from data that does not exist in that spreadsheet.
Building it with the people around the work
The crossing for the Ecosystem barrier is to build with the people who do the work, inside and outside the company.
Inside: the buyer, the shop and the project managers
Start with the purchasing lead, not the software. She knows which suppliers slip, in which season and on which items. Sit with her and the shop foreman and write down the questions they ask every week:
- Which open purchase orders are at risk for jobs shipping in the next three weeks?
- Which items do we buy from only one source, and what is the backup?
- When a supplier slips, which jobs and which customers does it touch?
The first AI job is to answer those questions from the purchase orders, job schedules and supplier confirmations the company already has, and to put the answer in front of the project managers before the delay reaches the site. For items the shop keeps on the shelf, a reorder point and safety stock calculation, built from each item's demand and its supplier's lead time, tells the buyer when to order and how much buffer to hold.
Outside: the suppliers themselves
The second step is the one most companies skip. Ask the five critical suppliers what they can share. Many already send order confirmations, ship notices and lead-time updates by email or through a portal. The galvanizer may be glad to send a weekly turnaround estimate if someone will read it. An assistant that reads those messages and updates the risk list does more than one that guesses from history.
This is where the supply chain becomes an ecosystem rather than a list of vendors. The fabricator and its suppliers start working from the same picture of what is due, and when.
Peers: worked examples you can check
Finally, look for people like you. Trade associations, regional manufacturing groups and suppliers who serve many shops are good places to ask who has tried what. A worked example from a shop your size, in your trade, is worth more than any vendor demo, because you can call them and ask what went wrong.
What changes when it works
For the fabricator in the example, success is not a dashboard. It is three changes the team can see:
- The project managers hear about at-risk material while there is still time to resequence the shop.
- Every single-source item has a named backup, reviewed once a quarter with the buyer.
- The weekly supplier update is read and acted on, because it feeds a list the production meeting already uses.
Each change rests on a person inside the company and a supplier outside it doing their part, with AI doing the reading, matching and flagging in between.
How Watch Tower fits
Watch Tower is Applied Intelligence for your business. Vulcan holds the people, assets, processes and evidence of an organization as connected objects, so a slipped delivery shows which jobs and customers it touches. The Console is one place where leaders see the business, its evidence and its decisions. Atlas, Watch Tower's supplier directory, connects equipment, materials and field services to the work. If you supply industrial businesses, you can learn about joining it on our supply partners page.
For more on where AI fits across your operation, read AI for traditional industries or AI for trucking companies, which covers the carriers in the middle of every supply chain.
Your next step: list your five most critical suppliers and the one question you wish you could answer about each every Monday. Then bring that list to us, and we will map it with you and the people who do the buying.
Frequently asked questions
How is AI used in supply chain management?
In a supply chain, AI does the reading, matching and flagging between people: it checks open purchase orders against job schedules and supplier confirmations, and it flags which orders put which jobs at risk. People still make the calls. The buyer, the project managers and the suppliers act on the list.
Are smaller companies using AI in their supply chains?
Most smaller U.S. businesses are still early with AI in the supply chain. Census Bureau survey data shows AI use is higher at large firms than small ones, so if you haven't started in purchasing, you are in the majority. Few worked examples exist nearby yet, which is why building with your suppliers matters.
Should I involve suppliers when using AI for supplier management?
Yes, involve your critical suppliers from the start, because they hold the information that predicts delays. Many already send order confirmations, ship notices and lead-time updates. An assistant that reads those messages and updates your risk list does more than one that guesses from purchasing history alone.
What data do you need to start using AI for operations and purchasing?
You can start with data most companies already have: open purchase orders, job schedules and supplier confirmations. Begin with the weekly questions your buyer already asks, such as which orders are at risk for jobs shipping soon and which items have only one source, and answer those first.
