Based on a conversation between Ricky Cecchini (VP of Product, CloudRadial) and Mike Parfitt (CEO, Team Metalogic) on ServiceAI, CloudRadial's transformative AI solution that solves tickets, scales technician expertise, and analyzes service desk performance.
When Mike Parfitt's UK-based MSP, Team Metalogic, became the first customer to implement ServiceAI, they weren't chasing the hype around autonomous AI agents.
They had a more fundamental problem: how do you maintain service desk quality when you're dealing with 30,000 tickets across multiple years?
"The reality is that you wish tickets had a full explanation all the time, and that's just not true," Ricky explains. "Most of the time what I see is a lot of tickets that just say, ‘fixed printer’”.
"The perfect ticket rarely exists," Mike adds. "I'd be surprised if I find a ticket that ServiceAI gives a 10 out of 10, just because there's always a different way of doing things."
That’s why ServiceAI addresses consistency and auditability.
Before diving into what ServiceAI does, it's worth addressing what it doesn't do: replace your technicians.
"I’ve yet to see a tool that can actually replace a technician," Ricky says. "ServiceAI is really meant to augment and help you find holes in your service desk so you can fix them."
That distinction matters. Too many MSPs approach AI with either unrealistic expectations (it will solve everything) or fears (it will eliminate jobs). The reality is more nuanced.
Let’s go back to the “fixed printer” example. If another technician (or the same technician six months from now) encounters a similar problem, that ticket provides zero value.
What did the first technician fix? Was it a driver issue? Hardware failure? User error?
"If they didn't document it, it never happened," Mike explains. "And these things very often come back in six months' time, and you absolutely cannot rely on the auditability of that ticket."
ServiceAI addresses this by scoring ticket quality on a 1-10 scale across three dimensions:
The key insight? This isn't about chastising technicians. It's about understanding where coaching is needed and measuring improvement over time.
Every MSP owner knows the documentation problem: strong intentions at client onboarding, then gradual decay as changes happen and nobody updates the docs.
"You onboard a new customer and you’re turning over all of those rocks; taking photographs, documenting everything that you see," Mike describes. "But then change management happens and configurations get adjusted, and whilst it might be in the ticket—is it also then replicated in documentation? The answer is rarely yes."
ServiceAI integrates with documentation platforms like ITGlue and Hudu to:
When a technician works on a ticket for Company A, ServiceAI references both general knowledge and Company A's specific documentation. Switch to Company B's ticket, and Company A's information is no longer accessible—preventing information bleed across clients.
Here's where AI's pattern recognition capabilities shine.
Consider a scenario where 20% of a client's tickets relate to performance issues in their virtual desktop environment. Individually, each ticket might seem like a one-off problem. Collectively, they point to a resourcing or scaling issue that the MSP should proactively address.
ServiceAI analyzes ticket data to identify:
"To be able to synthesize that five tickets are consistently password issues, or that there are hardware failures—it's not Dell, it's not Lenovo, it's not HP—it's hardware failures," Ricky explains. "You can take that kind of information to a client and it changes the conversation from 'Hey, you might need a couple replacements' to 'You have critical infrastructure issues that need to be addressed.'"
Traditional performance review approaches don't scale. Mike describes their old method: "We used to have a service manager that would randomly pick 5, 10 tickets a week to listen to call recordings and then compare that with the tickets. But it's so inefficient."
Spot-checking 5% of tickets randomly tells you almost nothing. You might pick the best examples or the worst examples purely by chance.
ServiceAI provides objective performance metrics across every ticket, identifying:
Importantly, this isn't about creating a surveillance state. "ServiceAI is useful as a way to actually figure out who needs assistance: how can you empirically find out who's doing the right stuff and who needs some coaching or correction," Ricky clarifies.
Mike shares several practical insights from Team Metalogic's implementation:
On data volume and relevance: They pulled in two years of tickets (approximately 30,000) because that represented their current service delivery model. Going back further would include outdated on-premises infrastructure that no longer reflected their cloud-first approach.
On the value proposition: "When we actually look at that data, and we see maybe 20% of tickets for a customer are performance-related issues when they're using AVD infrastructure—we can see that wider picture and understand that there's a resourcing or scaling issue at play here that is on us to remedy."
On practical concerns: Security matters. ServiceAI automatically scrubs personally identifiable information (PII) including passwords, email addresses, and phone numbers from training data—even though ITGlue integration requires password data access to pull article content.
Perhaps the most compelling use case goes beyond day-to-day ticket handling to strategic client conversations.
Traditional business review reports focus on SLAs, CSAT scores, and ticket type/subtype breakdowns. These metrics have value, but they're often generic and don't drive meaningful conversations about improving the client's operations.
Root cause analysis changes the game. Instead of saying "here's how many tickets you had," you can say "45% of your tickets stem from this specific accounting software—let's discuss training options or alternative solutions."
This transforms the MSP from a reactive support provider into a proactive strategic partner.
If you're evaluating AI tools for your MSP, Team Metalogic's experience offers several key lessons:
Transformation isn't about dramatic overnight change. It's about measurement, incremental improvement, and building systems that scale as your MSP grows.
For MSPs willing to approach AI with realistic expectations and a commitment to improving their foundational data quality, the tools exist today to start that journey. The future of autonomous AI agents remains exciting, but the present value of better visibility, documentation, and strategic insights is available right now.
ABOUT SERVICEAI
Discover how ServiceAI: