Praesto Troubleshoots With Mingo While the Client Is Still on the Phone

Matthias Kittok owns Praesto, an MSP, and was an OpenFrame beta tester. At the interview in June 2026, he ran OpenFrame side by side with his established RMM and installed the agent wherever he wanted Mingo's help.

His workflow fits a support call. When a client calls, he gives Mingo, OpenFrame's AI assistant for technicians, the machine name and a description of the problem. Mingo starts running while he's still talking to the client.

On one client's stubborn driver issues, Mingo dug in as deep within five to 10 minutes as he estimates would have taken him "probably days". His take on the product so far: "It saves me a lot of troubleshooting time already."

Praesto is a small shop. Matthias puts it "somewhere in between" a very small MSP and a team of five to 10 techs. Clients call him directly when something breaks.

Tricky issues take time. To confirm a hunch, he'd have to write a script or do something else to check it. If he can't solve it on the call, the client hears "I'll call you back later" or "I'll get back to you before the end of the week".

He's also wary of the "single pane of glass" pitch. His test is simpler: does everything talk to each other "in how I want to run my business"?

Matthias joined OpenFrame as a beta tester and runs it next to his established RMM. He installs the agent on the devices where he wants Mingo.

The workflow is built around a support call: "Usually how I started is I just give the machine name to Mingo and I say this is the description of the problem. And it just starts running while I'm talking to the client."

While Mingo works, he thinks through his own experience on what else it could be. Mingo works alongside that judgment and gets him there faster: "Often the AI aligns with what my experience has been. It just gets there quicker or gets to a solution quicker than I could because I'd have to write a script or do something to confirm my suspicions."

He credits the depth to osquery, which runs under OpenFrame: "it knows all. It can tell everything that's going on in that machine." He also likes the product's direction. Flamingo is the first vendor he's seen "dive full in" and build the product around the AI instead of plugging the AI in later. His own caveat: "Whether that's technically the reality or not, that's kind of the way it seems."

At the interview in June 2026, Matthias hadn't tracked hours saved. His results are about speed on tricky issues. The driver example is his own estimate for one client issue: five to 10 minutes for Mingo to dig as deep as he thinks would have taken him "probably days".

The goal he cares about most is first call resolution: "So first call resolution, I think, is the kicker here." If he can work the problem with the client still on the phone, he doesn't have to promise a callback. In his words: "That's huge."

His rollout grew from there. When he ran into one-off issues on devices without the agent, he deployed it: "I started doing that more and more." He has also sent the Flamingo website to multiple people.

What he still wants, framed as polish: monitoring policies with a hierarchy (global plus per client), grouping and tagging, custom fields that scripts can use as variables, patch scheduling with maintenance windows, third-party patching, a link from vulnerabilities to remediation, a branded icon for Fae (OpenFrame's agent on client machines), per-client AI guardrails, Fae in Slack or Teams for tickets and approvals, org-level dashboard views, and easier onboarding from another RMM.