Embedding New AI Features for Hardware Already in the Field
Thread AI Editorial · July 31, 2026
A device maker sells fax and multifunction devices to businesses that still receive orders, referrals, and records on paper.
With Lemma, Workers read and classify each page the moment it lands and file it in the customer’s own systems. The device maker sells the Worker as a new feature for hardware its customers already own.
What Lemma delivers
New revenue stream unlocked
From hardware already installed.
Scalable to new customers
By connecting each one’s systems through configuration.
AI features customers are asking for
Delivered on devices they already own.
The Challenge
In sectors like manufacturing, many referrals, orders, and records still arrive by fax. The device captures the page, but someone at the customer’s site still has to read it, decide where it goes, and type the details into another system. It’s slow, repetitive work that happens for every page.
The device maker’s hardware already sits at the start of that process. But adding AI to devices in the field and connecting them to each customer’s own systems would normally mean custom engineering for every customer.
The Solution
The device maker’s team builds Workers on Lemma and sells each one as a new feature for devices its customers already own. When a page arrives, the Worker classifies it and acts on it. A faxed order, for example, can open a purchase order in the customer’s ERP and message the account owner that it’s there. Any action that writes to a customer’s system waits for a person to approve it.
Customers stop retyping faxes, and the device maker has a new product to sell on hardware that’s already installed.
Lemma Capabilities
Classifies pages on arrival.
Referrals, orders, and everything else are sorted the moment the device captures them.
Reads and writes back to each customer’s systems.
Each customer’s system of record or ERP is connected once, so every new integration is configuration work for the device maker.
Embeds new AI features on hardware already in the field.
The device maker can add AI features to devices customers already own, without replacing the hardware.
A person for every uncertain page.
Anything below the customer’s confidence threshold goes to a review form for human approval.
What’s Next
With every page classified and filed as it arrives, staff at customer sites can act on orders and referrals as soon as they land, and the device becomes the first step in each customer’s workflow. For the device maker, each new Worker it builds is another feature it can offer across the hardware already in the field.
Does this sound familiar?
Your product is installed everywhere, but every new feature means custom work for each customer.
You want to sell AI features without asking customers to replace what they already own.
Your customers receive important documents in formats their systems can’t read.