Thread Voice and Service Intelligence Revolutionize the MSP Service Desk
Thread Voice AI and Service Intelligence are revolutionizing MSP service desks by enhancing phone support efficiency and knowledge management.
Jason Sola has spent 26 years in IT, 13 of them inside MSPs of every size. A few months ago, he and five colleagues walked away from that model and started a new company from scratch and called it Solon Intelligent Services.
Its Tier 1 service desk runs entirely on AI. Every password reset, every printer driver, every "I can't get into my email" goes to AI first. Clients know that going in. Jason tells them on the first call. And so far, the companies he wants to work with like hearing it.
That's a big bet for a business about two and a half months old. It's also a preview of where the AI service desk for MSPs is headed, and a lesson for everyone who didn't get to start with a blank page.
Anyone who's worked in an MSP knows the feeling. The team is running at 120, 130% of capacity. Tickets keep coming. Everyone agrees the answer is another body. Then leadership says the budget isn't there, or the hire takes four months, or the new tech burns out and leaves before they're up to speed.
Jason lived that cycle for over a decade, so he flipped the order. At Solon, the core managed services stack (RMM, patching, monitoring) is table stakes. The difference is that AI and automation absorb the volume, so a small team can move fast without throwing bodies at every spike. That's what MSP scale without hiring looks like in practice. Growth stops automatically meaning another open req.
Solon packages its service per user, all in, with unlimited AI support. Tier 1 is handled by AI, backed by Thread. Human escalations come from a pooled bucket of hours, roughly half an hour per person per month across the client organization, with a standard rate beyond that.
The pooled hours are deliberate. Jason wants clients to actually work with the AI instead of typing "escalate" the second a chat opens. He's upfront that both sides are still learning where the limits sit. But the structure rewards the behavior that makes the model work.
"Right now we're just going to hire into tier 2 and let AI do the tier one work," says Jason Sola.
Think about what that removes. The classic Tier 1 job is password resets and printer drivers. New hires want out as fast as possible, so they promote quickly, and the MSP is stuck in a permanent loop of interviewing and backfilling. Hiring straight into Tier 2 breaks the loop. People start on more interesting work. They stay at that level longer before they're ready for Tier 3. The business gets more longevity out of every hire.
Jason has gotten pushback on this. Can an entry-level person really handle Tier 2 issues? They can when the service desk feeds them context. When Thread flags that an issue has been seen before and shows how it got solved, a newer tech has guidance right inside the workflow. Jason jokes that interview candidates used to say, "I don't know, but I have Google." Now it's, "I don't know, but I have Super Magic."
That's the real promise of MSP Tier 1 automation. People get better first jobs.
Ask any tech what they'd happily never do again and you'll hear the same list. Time entry. Deduplicating tickets. Digging through the catch-all inbox to sort spam from auto-forwards from replies to a booking link.
Time entry is the classic example. It's critical. Without accurate time, you don't know which ticket types eat your margin, which clients are profitable, or where you're missing billable work. But the engineer who'd rather spend those 15 minutes on the next issue has a point. Jason wants that engineer on the next issue too. So time entry gets automated. An on-site tech can talk into their phone ("Bob tapped me to fix the printer, took 15 minutes, updated the drivers") and the ticket, time entry, and client update all get handled.
Same with the catch-all. At Jason's last MSP, letting AI read, classify, and close junk tickets was one of the first wins. Nobody touches those tickets anymore. Nobody misses it.
Each of these is small on its own. Stacked together, they're how Solon keeps its technician-to-endpoint ratio high. Pair the service desk with AI-backed tools elsewhere in the stack (Solon uses Lexful for documentation) and each engineer covers far more ground without getting buried.
The fear with AI on the front line is that clients will hate it. Jason's view comes from his own experience as a customer.
"I don't care who gives me that answer as long as I get the answer I need and I can get on with the work that I'm doing," says Jason Sola.
Every year, he calls his satellite radio provider to cancel so a human will offer him a better price. This year, a voice agent picked up and offered a better monthly rate, locked in. No human had ever offered that. It was the best renewal he's had.
A bad bot is a nightmare. A good one that solves the right problem right away is about the best service there is. That's the bar. It's also why Jason was most excited when identity verification landed in Thread. He wants AI handling password resets, but never without confirming who's on the other end. With verification built in, the agent confirms the user and finishes the job. The most common Tier 1 request closes without a human.
His north star is the zero-touch ticket. The client gets the outcome and nobody on the team ever opens it. Thread's AutoMagically is built for exactly that. An agent figures out what needs to happen, then runs a deterministic runbook to do the work the same way every time. One MSP using it has already reached 17% of tickets closed end to end. Jason wants to push that number as high as it can go, effectively and securely.
Solon's AI-first service desk does a second job. It sells.
Jason's clients want to run leaner too. They want AI in their own businesses, and they want to adopt it securely. When the provider's own support is fast and clearly AI-powered, that conversation starts itself. If your clients still have to send an email and wait for a callback, nobody is calling you to help them build an agent.
When that work does come, price it like a service. Automations need care and feeding. Tools change. You learn more. Sometimes you rebuild the whole thing a better way a year later. Jason's take: charge for it as recurring revenue. A one-time $5,000 project doesn't build a practice or help your valuation. Recurring revenue funds the AI engineer who goes and finds the next problem worth solving.
Most MSPs aren't starting from zero. You have clients paying real money for the service you deliver today. Flipping everyone to an AI-first desk overnight carries real risk, and being careful is fair.
Jason has two suggestions.
If you have the resources, spin up a separate AI-first team. Go after the clients Solon is going after: companies that want to be AI-first and security-first. You'll probably find a few existing clients who want to move over too.
If you don't, start in bite-sized chunks. Turn on automated ticket triage. Jason says he'll never hire a dispatcher because Thread handles that for him. Clean up the catch-all. Automate time entry. Bank the wins, find the next efficiency, and keep chipping away until you get to a fully agentic service desk.
The destination is the same either way. The only real choice is how soon you start.
Jason left a long MSP career to bet on this model because he thinks the shift is inevitable. You don't need to make the same leap to get the same gains. Start with triage, verification, and the work your team hates most. Build from there. And if you want to hear how other MSPs are doing it, the AISU (AI Service Unleashed) community meets every week.
Ready to take Tier 1 off your team's plate? Book a demo and we'll show you where to start.
Thread Voice AI and Service Intelligence are revolutionizing MSP service desks by enhancing phone support efficiency and knowledge management.
Most MSP service desk automation fails because it's bolted onto a broken system. Here's the three-layer approach that actually changes service...
Discover how Model Context Protocols (MCPs) empower MSPs to enhance AI integration, streamline operations, and improve service delivery across their...
Sign up for our newsletter to get the latest and greatest delivered directly to your inbox by Aerosmith. just kidding, it's us, and the newsletter is awesome.