Solo MSP Cleans Up Hundreds of Endpoints Across a 1,290-Device Fleet

Jay Contor runs Our IT Guyz alone. 1,290 endpoints across Australia, one technician, no team behind him. The typical MSP ratio is 100 endpoints per tech. Jay is at nearly 13 times that, and he gets fewer than 20 tickets a day because he has systematized everything he can. What he can't systematize is the proactive work: the cleanups, the SSDs filling up, the users whose three-screen setups on four-year-old laptops are going slow. That work would require a road trip across Australia or another technician, and Jay isn't hiring. He started testing OpenFrame to find out whether AI could pick up the proactive work he never has time for. Within a few days, he'd deployed the agent across his entire fleet and run fleet-wide disk cleanups that his RMM couldn't touch.

Jay moved to Ninja for RMM and Syncro for ticketing because both were simple enough to run without a team. He trained his clients to log tickets through a tray icon so he didn't have to. He evaluated Halo PSA for two years and walked away because it couldn't support the tray-icon workflow his clients already knew. What he couldn't cut was the proactive maintenance layer. Dozens of machines across his fleet were running out of disk space. Some of those clients were hours away, and the options were either a road trip of SSD swaps or waiting until the machines failed hard enough to justify a hardware replacement. Neither option scaled.

Jay started small. A handful of machines, just to see what OpenFrame's AI could do. The first cleanup freed more disk space than his existing tooling had been able to touch. That was enough to convince him to scale. He asked Mingo to run the same cleanup across every non-server device in his fleet, and Mingo worked through them one by one, asking for script approvals along the way. Hundreds of machines cleaned up without a plane ticket. The printer install was a separate test. A portable printer next to him, a laptop that had never seen the driver, and a plain-English request to Mingo. It took longer than installing it manually, but the driver went on. Beyond the one-off tests, what Jay saw in Mingo was a faster path to the information every RMM already has but makes you hunt for. Ask which processes are slowing a machine down, and Mingo answers directly instead of making a technician dig through three screens of telemetry.

Fleet-wide cleanups Jay had been putting off, the kind that would have meant either a road trip of SSD swaps or waiting for machines to fail hard, got done without him touching each device. Ninja tells him which machines are running out of space. OpenFrame actually clears the space. On diagnostics, a slow-machine ticket that used to mean jumping into the RMM and digging through telemetry is now a plain-English question to Mingo. The answer comes back in seconds. The time savings are still early-stage, but the direction is clear: the proactive maintenance layer Jay never had hours for is now work the AI handles while he focuses on the tickets only a human can solve. At $5 per device per month, Jay has already said he'd pay, even if OpenFrame did nothing beyond this.