Rehabmart, an online retailer of rehabilitation and medical equipment, runs Makez.ai AI agents across purchase orders, product content, catalog data entry, order tracking, and quotes. Here is what that looks like inside Rehabmart's own workspaces.
Rehabmart was founded in 1998 by Mike Price, OTR, and Hulet Smith, OTR/L, two occupational therapists. That clinical background still shapes the catalog: patient lifts, mobility aids, pediatric equipment, fall prevention products, and supplies for senior care, schools, and the Veterans Affairs market.
A catalog and order book that wide means constant paperwork: supplier purchase orders in different formats, new SKUs to enter, and shipments to track. That is the operational reality behind most growing retail and eCommerce businesses, and it is where Rehabmart brought in Makez.ai.
None of this was a people problem. Rehabmart's team knew the products and the customers. What they didn't have enough of was time, because so much of the day went into reading, typing, and re-typing the same information.
Every supplier PO and invoice had to be opened and keyed in by hand. See how the Document Agent changed this.
New SKUs arrived faster than product pages could be written. The Content Agent now drafts the first pass.
SKUs and attributes typed in one field at a time, now handled by the Catalog Agent.
Tracking lived in emails and spreadsheets. Now it's the Order Tracking Agent's job.
Details drifted apart until someone noticed. The Contact Agent is closing that gap.
Pricing lookups and formatting took hours from selling. Quote automation now runs through the Workflow Copilot.
Each tab below is a real screen from Rehabmart's own Makez.ai account. Reviewer names and a few identifying fields are blurred, everything else is exactly as the team sees it.
Supplier spec sheets go in, and the Content Agent drafts the page copy, ready for a human editor to check before it goes live.
This workspace also feeds Rehabmart's buying guides, the same kind of reviewed writing covered in the Makez.ai research and guides library.
Reviewer names blurred for confidentiality.
The Catalog Agent reads supplier files, fills in SKU fields, and routes anything uncertain to review rather than guessing.
Records stay linked to their source file through the built in document management and CRM layer.
Reviewer names and SKU details blurred for confidentiality.
The Order Tracking Agent matches incoming files to tracking numbers and flags anything with a missing ID, the kind of real time monitoring covered in AI agents for operations.
At this volume, exceptions are the real work. Surfacing a problem early beats processing easy cases quickly.
Reviewer names and tracking numbers blurred for confidentiality.
Rehabmart's most mature workspace. Purchase orders and invoices, including templates for its Veterans Affairs business, are read by the Document Agent and matched against open orders using logic similar to the Reconciliation Agent.
Of 27,167 documents, 25,249 have been processed without anyone opening them by hand, about 93 percent. The rest sit in a review queue, exactly where a person should be looking.
Reviewer names and document IDs blurred for confidentiality.
Rehabmart's next workspace automates quote generation: pricing lookups, margin checks, and proposal formatting, handled by the Workflow Copilot so sales reviews and sends rather than building from scratch.
Order processing is also expanding beyond purchase orders, pairing the Document Agent with the Reconciliation Agent and the Insights Agent for a clearer read on order volume. Everything new is built in Agent Studio and staged through Testing & Release before it touches a live account.
An early preview of the redesigned workspace, with sample data.
Every workspace reads from and writes back to the systems Rehabmart already runs, connected through the Integration Fabric and given shared memory through the Business Context Engine.
Monitored under Rehabmart's trust, security, and governance settings, built in Agent Studio and shipped through Testing & Release.
These figures come directly from the workspaces above, not projections.
This is still early in Rehabmart's automation journey. Quotes and deeper order processing automation are next, following the same pattern used so far: prove one workspace against real volume, then move to the next. More in the Makez.ai customer stories library.
A short conversation about bringing AI agents into a working eCommerce operation.
The Integration Fabric connects to systems already in place.
The Business Context Engine gives every agent shared memory.
Uncertain cases route to review rather than getting guessed at.
Describe the workflow, Makez.ai builds it. See Build in Plain English.
Every agent is staged through Testing & Release first.
Audit trails and access controls, covered under Trust, Security & Governance.
Rehabmart is one story in a wider pattern. Other industries automate the same categories of work, each with a different starting point.
Read more in the Makez.ai customer stories library, or explore AI agents for business teams generally.
Where the busywork usually shows up first, and which agent handles it. Quotes and order processing, the two Rehabmart is expanding next, are marked below.
| Agent | What it takes off your plate | Best starting point for |
|---|---|---|
| Document Agent | Reads and extracts data from purchase orders, invoices, and forms. | Order processing, teams still typing supplier paperwork. |
| Reconciliation Agent | Matches transactions and closes the loop on every order. | Order processing and finance close. |
| Workflow Copilot | Pricing lookups, margin checks, and proposal formatting. | Quote automation for sales teams. |
| Catalog Agent | Keeps product catalogs accurate across channels. | Fast growing SKU counts. |
| Content Agent | Drafts product copy for human review. | Backlogged product pages. |
| Order Tracking Agent | Tracks orders and flags exceptions early. | Customer teams fielding status questions. |
| Contact Agent | Cleans and deduplicates vendor and customer records. | CRM data that has drifted out of sync. |
| Insights Agent | Turns document and order volume into a readable trend. | Leaders wanting a clearer read on the business. |
Closer to finance or operations than retail? Start with AI agents for finance and back office teams or AI agents for operations teams instead.
An online retailer of rehabilitation and medical supplies, founded in 1998 by two occupational therapists, Mike Price and Hulet Smith. It sells mobility aids, patient lifts, pediatric equipment, and other therapeutic devices.
Six or more, including purchase order and invoice processing, order processing, product content, catalog data entry, order tracking, vendor record cleanup, and quote automation, with more rolling out over time.
Yes. Order processing runs through the Document Agent and Reconciliation Agent. Quote automation runs through the Workflow Copilot.
No. Agents connect to what's already running through Integration Fabric, no rip and replace required.
Yes. Agents stage data for review, and a person confirms it before submission. More on how this is governed under Trust, Security & Governance.
Rehabmart started with one workspace. Talk to Makez.ai about where the busywork is heaviest in your operation, and what to automate first.