Mear Technology Finds OpenFrame's AI 'Absolutely Phenomenal'

Stephen Mears runs Mear Technology, a Scottish MSP established in 2011. It has five technicians and about nine staff, and looks after about 1,000 endpoints and 1,500 users. Its clients are small and medium-sized businesses, mostly across Scotland's central belt, plus some international clients.

Stephen tests RMM platforms regularly but has never switched, because none of them offered enough to make a move worth it. He tested OpenFrame in beta alongside his existing RMM, which stayed his main platform. He rolled it out with a single script from that RMM, and some of his end clients have used Fae, the AI agent that sits on their machines.

At the interview in March 2026, the AI was what stood out. He can type one chat message asking it to find the Windows 10 machines across his infrastructure and upgrade them to Windows 11, and it goes off and does it. Jobs that can take five minutes to an hour in his current RMM run at once. And one fix he'd have spent a week on, the AI solved. In his words: "What you have done with the AI is absolutely phenomenal."

Mear Technology runs a mature stack. Stephen started out as a one-man band and has been doing this for 14 years. Today the business runs an in-house PSA, a large script library in its RMM, and N8N automations built on that RMM's APIs. One of those workflows advises clients on Windows 10 end of life.

He evaluates RMM platforms often. Each one has its pros and cons, and none has made switching worth it: "The real problem is that none of them really offer significant value in terms of switching."

Day to day, the friction is time. In his current RMM, a quick job sends a single script to a device. Sometimes it takes five minutes, sometimes 10 minutes, sometimes an hour to run. When one of his team calls about a slow machine, someone has to find that machine and connect to it before any diagnosis starts.

Stephen tested OpenFrame in beta alongside his existing RMM. Rollout took one script. He built a script in his RMM that deploys OpenFrame, using a variable set on each client site, and pushed it out from there.

Most of his use runs through chat with the AI. He can ask it to look across his infrastructure, tell him how many Windows 10 machines there are, and try to upgrade them to Windows 11. In his words, "it goes off and does it."

Diagnostics work the same way. When one of his team calls about a slow machine, he can ask the AI to find that machine and run a diagnostic on it, without connecting to it himself: "Your AI is significantly better because I don't even have to connect to the machine."

Fae is the client-facing piece. It sits on end-client machines, and some of his clients have used it.

At the interview in March 2026, the results were about speed and reach, not measured numbers. Stephen still ran his existing RMM as his main platform, with OpenFrame alongside it.

Speed came first: "What's really cool about yours is I asked AI to do something and it's an instant. It's just so quick. It's awesome." Jobs that can wait five minutes to an hour in his current RMM run at once.

Fleet-wide work by chat is where the time goes back. Asking the AI to find and upgrade Windows 10 machines across his clients is, in his words, "saving us hours of time."

One fix stood out. "Some of the fixes it gave to us was just awe-inspiring," he said. One problem the AI solved would have taken him a week.

His summary of the beta: "So I guess we love OpenFrame. We think it's awesome." And: "It's been the most captivating product we played with for a long time."

What he still wants, framed as polish: custom fields at device, site and global level to record script results; richer device lists, saved filters and fast search; script scheduling and monitors; AI actions logged as per-client tickets; brandable Fae with technician alerts and two-way chat; patch management, plus device details such as the last logged-in user; and integrations with existing PSAs and tools, rather than replacements for them. The scheduling ask ties to his own idea for what comes next: "We should be using the AI to build more autonomous automated processes."