AI is the hottest topic in tech right now. Every vendor has an “AI-powered” feature, every conference session has an AI angle, and every MSP wants to know how AI fits into their business.
But here’s the truth: AI isn’t magic. At least not yet. As Tim Barton-Wines from HaloPSA put it during our recent Talking VoIP with Datagate podcast, most of what’s being branded as AI today is really just smart data pattern recognition—combined with a hefty license fee if you’re unlucky enough to buy from the wrong vendor.
That honesty is refreshing. Too often, MSPs are told they need AI for the sake of appearances. But if you cut through the hype, you’ll see that AI already has practical, tangible applications that can help MSPs reduce costs, improve resolution times, and boost profitability. The trick is knowing where to apply it—and where not to.
This post highlights key takeaways from the conversation with Tim Barton-Wines and places them in context with broader trends in SaaS and telecom billing.
AI’s Real Strength Today: Pattern Recognition and Suggestions
Tim was clear about where HaloPSA sees the most success with AI: analyzing service data and spotting patterns that humans would miss.
Take ticketing, for example. HaloPSA has vectorized its database so that when a new ticket arrives, the system looks at the subject, body, keywords, sentiment, and context. It then automatically compares that ticket to thousands of previous cases and pulls back the most similar examples.
From there, AI can:
- Suggest a likely category.
- Propose a time target based on historical fixes.
- Pre-fill scripts or responses that solved similar issues before.
That’s not futuristic science fiction. It’s practical automation that saves technicians hours every week. And in our industry, hours saved directly impact profitability.
As Tim put it, this isn’t about replacing technicians—it’s about enriching their work. Imagine a junior tech picking up a ticket and already having the top five likely solutions in front of them. That’s efficiency. That’s margin protection.
The Right Role for AI: Your Graduate Trainee
One of Tim’s best analogies was to treat AI like a graduate trainee.
Would you give a brand-new trainee the keys to respond directly to clients, without oversight? Of course not. The same applies to AI. Today’s systems are good at augmenting decisions, but they’re not at the point where they should be left unsupervised.
That’s why HaloPSA defaults to “suggest mode.” AI will recommend a categorization or action, but the technician has the final say. That safeguard not only preserves accountability, it builds client trust.
As MSP leaders, we can’t afford “lazy approvals.” One wrong automated response could cost a client relationship. AI should be a tool that supports technicians, not a substitute for them.
What MSPs Really Want from AI
Interestingly, when asked what MSPs are requesting most, Tim said it’s not specific features. It’s simply that they “want to be seen as using AI.” Clients, vendors, and even peers expect it.
That reflects where we are in the adoption curve: early, experimental, and driven by perception as much as results. The job of a good PSA or telecom billing platform vendor is to take advanced AI capability and make it accessible and safe for MSPs who may not have in-house expertise.
This aligns with what we hear at Datagate too. MSPs aren’t asking for AI to run invoices on their behalf. They’re asking for insights: “What should I know before I send this billing run? What’s unusual about this month compared to last month?”.
That kind of virtual-CFO capability is exactly where AI can shine in telecom billing.
Transparency and Control Are Non-Negotiable
Many MSP owners often raise the same concern: “How do I trust AI decisions if I can’t see how it got there?”
Tim’s answer was spot on: transparency isn’t optional. In HaloPSA, every AI suggestion comes with a clear log, and the technician chooses whether to apply it. Users can even audit, override, or disconnect the AI module if they want more control.
That’s the only way to build trust. If your vendor can’t explain where AI results are coming from—or worse, if they’re just piping your tickets into ChatGPT without security guarantees—walk away.
Real-World Examples of AI Helping MSPs
Tim shared some fascinating examples of HaloPSA users applying AI today:
- Emotion detection on tickets: AI rates the sentiment of incoming emails and flags negative ones so managers can intervene quickly. That protects customer experience before small frustrations escalate.
- Automated root-cause analysis: By integrating HaloPSA with NinjaOne and Azure Runbooks, one MSP built a workflow where low-disk-space alerts automatically return file analysis and remediation suggestions. Technicians pick up the ticket with answers already in hand—saving 30+ minutes per case.
- Fighting social engineering: With AI-powered MFA workflows, suspicious calls can be verified before sensitive actions are taken. In an era of AI-generated voice scams, this is a crucial safeguard.
Each of these examples shows a principle: AI works best when it augments existing processes rather than replacing them.
Measuring AI’s Real Impact
If you’re starting to use AI in your MSP, don’t get lost in vanity metrics. Stick to the same KPIs you already track—resolution time, first-contact resolution, NPS—and add one more: the rate at which AI suggestions are accepted and applied.
If adoption is high and those suggestions correlate with faster service or better satisfaction, you know AI is delivering real value.
The Road Ahead: Agents, Bots, and Beyond
Looking forward, Tim sees AI in PSAs moving toward digital “agents” or bots—specialized workers trained for dispatching, triaging, or managing projects. HaloPSA is even exploring building its own AI engine in-house to reduce reliance on third parties.
The dream scenario for MSPs? A future where AI helps solve the talent shortage by scaling service without endlessly hiring. Lower costs, higher margins, and better customer experience—all at once.
That’s not here today. But it’s closer than many think.
Guidance for MSPs Considering AI
MSPs evaluating AI adoption should follow a measured approach:
- Get your processes right first. Don’t throw AI at a broken system. Fix your workflows, then let AI multiply their effectiveness.
- Start small, measure, and scale. Begin with safe, out-of-the-box features like ticket suggestions or billing run insights. Track results, then expand.
- Demand transparency. Only work with vendors who can show you what their AI is doing, where your data is going, and how you stay in control.
- Focus on enablement, not replacement. Use AI to free up technicians so they can work on automation and higher-value projects. That’s where the real efficiency gains come from.
Closing: Turning Hype into Advantage
The AI conversation in MSPs today is noisy, full of marketing claims and inflated promises. But if you cut through the hype, there are practical, revenue-impacting use cases right now.
Do your best to separate signal from noise. The winners in the next few years won’t be the MSPs who slap “AI-powered” on their website—it’ll be the ones who implement AI thoughtfully, safeguard accountability, and use the time they save to invest in automation and customer experience.
As Tim reminded us, AI is best treated as a smart trainee, not a replacement. Train it well, give it good processes to build on, and hold it accountable. Do that, and AI won’t just be another buzzword—it’ll be a genuine lever for growth in your MSP.
Learn about the Datagate-HaloPSA Integration
