Xyte.

The Company That Sells Hardware Its Second Act

Here is a fact about the hardware business that nobody puts on the box: the sale is the easy part. A manufacturer builds a projector, or a video bar, or a rack of audio gear, ships it, and then - in the traditional telling - the relationship ends at the loading dock. The customer owns a thing. The maker owns a warranty obligation and a hope that they buy again in five years. This is a perfectly fine way to run a business, in the way that selling someone a single meal is a fine way to run a restaurant. It is just not the way that produces recurring revenue, or a reason to open the dashboard every morning.

Xyte, founded in Tel Aviv in 2019, exists in the gap between "we shipped it" and "we still know how it's doing." It is a cloud platform - the company calls its flagship the Xyte Device Cloud, and elsewhere a Connected Device Management Platform, which are two names for the same idea - that lets a manufacturer connect its physical products to the internet, watch them, fix them, update them, and, crucially, bill for them on a subscription. The industry has a suitably clunky word for this: servitization. The plain version is: sell the outcome, not just the box.

The people who built it come at the problem from both sides. Omer Brookstein, the CEO, spent close to 15 years at Crestron Electronics, the pro-AV giant, where he helped stand up the company's Israel office and ran enterprise sales. He knows the loading-dock problem from the inside - he spent a career on the wrong side of it. His co-founder, CTO Boris Dinkevich, comes from the software world, where he built cloud applications for large tech firms. One founder who knows exactly how hardware companies think, and one who knows how to make software behave: it is a sensible division of labor for a company whose entire pitch is teaching hardware to act like software.

What you can actually do with it

Strip away the category jargon and Xyte does a handful of concrete things. It monitors fleets of devices in real time, so a manufacturer or an IT team can see, from one screen, which units are healthy and which are about to sulk. It pushes firmware updates centrally, instead of sending a technician to each site with a USB stick. It runs 24/7 health checks that catch faults before they become outages - the company says this cuts downtime meaningfully. And it wraps all of that in the commercial machinery - subscriptions, billing, even equipment financing - that a hardware company needs if it wants to sell its product as a service rather than as a slab.

The people who benefit are not only the manufacturers. System integrators - the firms that install and maintain AV and IT gear for offices, campuses, and hospitals - get a way to turn one-time installations into ongoing, recurring contracts, an arrangement the trade press has taken to calling AVaaS, or AV-as-a-Service. Managed service providers get a control plane. And the enterprises at the end of the chain get devices that tend to fix themselves before anyone notices they were broken.

The interoperability problem, and the edge problem

Two things make device management genuinely hard, and Xyte has shipped a product at each. The first is that the world is full of rival ecosystems that don't talk to each other - a room might have a Zoom bar, a QSC processor, and a Biamp box, each with its own cloud. Xyte's Connect+ platform and its cloud-to-cloud connectors, rolled out through 2025, aim to pull those separate clouds into a single pane of glass. Unifying rival vendors is harder than it sounds, because interoperability is a promise everyone makes in a keynote and few keep in a shipping product.

The second problem is age. A great deal of the world's installed hardware sits on a local network, was built before "cloud" was a verb, and was assumed to be un-cloudable. In early 2026 Xyte made Secure Edge generally available - an appliance that extends its cloud management to legacy, on-prem devices without ripping out and replacing what's already bolted to the wall. This is the unglamorous work of infrastructure: meeting the installed base where it is, rather than insisting it start over.

The AI part, told without the hype

In late 2025 Xyte announced what it calls its first AI Teammate, with general availability targeted for early 2026. The claim is narrower and more useful than most AI announcements: when a device throws an incident, the Teammate reads the logs, pulls the relevant manuals for that model, and walks a technician through the fix - or handles the first line of it outright. It runs on what the company calls a proprietary Semantic Context Graph rather than a generic chatbot, which is a fancy way of saying it is built to understand a specific device's context instead of guessing. The business case is unromantic and therefore believable: fewer support tickets, lower support cost, better uptime, no extra headcount.

Investors have found the whole thesis persuasive. In January 2024 Xyte announced a $30 million round - a $20 million Series A led by Intel Capital, with Samsung Next joining and existing backers S Capital and Mindset Ventures returning, plus $10 million in venture lending from funds managed by BlackRock. Intel Capital's Roi Bar-Kat took a board seat. It is a notable roster for a company of roughly 49 people managing, by its own account, millions of devices - which is either a sign of enormous leverage or a reminder that the best infrastructure businesses are quietly small.

The customer list is the tell. Xyte works with manufacturers across pro AV, unified communications, smart buildings, medical, robotics, and industrial - and the names that surface include Crestron, Legrand, Poly, QSC, Biamp, Zoom, and Panasonic. When a company's users are the brands you already recognize, and the company itself is one you'd never heard of until now, that is usually the signature of infrastructure doing its job: invisible, load-bearing, and quietly everywhere.