Challenge
System Protection Partners handles roughly 2,600 tickets a month with a service team of about 10. Ticket classification was often blank, dispatch relied heavily on load, and years of documentation were difficult to retrieve quickly. Technicians often escalated when the right prior resolution was buried in the knowledge base or ticket history.
Solution
ConnectWise AI Agents™ were deployed directly inside ConnectWise PSA™. Summary and classification agents set the company, summary line, type, and subtype at intake. The dispatch agent follows written SOP conditions and group logic. In-platform AI search pulls KB articles and prior resolutions, including similar tickets from other clients, into the technician’s view.
Results
CW AI Agents now touch every ticket on the help desk board. Ticket data is consistent enough to report on for the first time. Dispatch is more granular and SOP-driven. Search brings documentation and prior fixes to Tier 1 in seconds, giving frontline technicians more Tier 2-level context. James estimates at least five minutes saved on research-heavy tickets (directional, not formally measured).
From blank fields to intake-ready tickets
System Protection Partners has grown alongside the ConnectWise ecosystem for close to a decade. Today, the company averages 2,600 helpdesk tickets a month which is staffed by a team of 10 across Tier 1, Tier 2, and engineering roles.
Before CW AI Agents, the help desk intake process was thin by necessity. Ticket types and subtypes were cumbersome to set. With 10 or 15 types and just as many subtypes, technicians often could not tell which choice was right, so the fields were often left blank. Now the agents classify every ticket at intake. When the output is missed, the team revises the criteria in the SOP and improves the next result.
[ConncetWise] AI Agents touch every help desk ticket. Types and subtypes were hard for our technicians to manage, so we often were not capturing them at all. Now the agents do that work, and we can finally see where issues are landing without asking technicians to spend time on it.
The enrichment goes beyond classification. When James emails a ticket in on behalf of an executive, the agents read the subject line and body. They set the right company and summary automatically. That prevents tickets from being misfiled under System Protection Partners’ own company record and leaves technicians with less cleanup to do.
Dispatch guided by written SOPs
Dispatching used to run largely on load, and the team was not always confident the work was landing in the right place. With CW AI Agents, routing is defined in plain language. The conditions are written as SOPs, and tickets can be appended to specific groups rather than assigned only to whoever is next.
Search that makes existing knowledge usable
The knowledge was never the problem. System Protection Partners has a standing rule: every resolution that requires research should end with a KB article. Years of that discipline live in documentation. The challenge was retrieval. Technicians had to guess the keyword or title the original author used before the right answer appeared.
The documentation was already there. The hard part was finding it. [ConnectWise] AI Agents read the full article and ticket history, then put the right information in front of the technician in seconds. That’s where the value is.
The agents surface similar tickets from other customer accounts, not just the one in front of the technician. That gives technicians context they could not reliably find through a keyword search, and it makes the time-saving practical rather than theoretical.
The strongest outcome: Tier 2-level capability at Tier 1
That search capability changes who can close the work. An issue a senior technician solved last quarter can become a documented fix that appears for Tier 1 the next time it shows up. The result doesn’t replace Tier 2; it gives frontline technicians more of the context they need to resolve issues themselves before escalating.
It's not replacing anyone. The work is still human-driven. What changes is that Tier 1 can get closer to Tier 2 capability because the agent brings the fix, prior ticket, or KB article to them at the right moment.
James considers at least five minutes saved per ticket a conservative estimate, while being clear that the team has not formally measured it. The savings are most noticeable on research-heavy tickets, where a faster search can determine whether a technician resolves the issue directly or needs to escalate.
AI inside the system where technicians work
System Protection Partners describes itself as a late bloomer on AI. The team started cautiously, using Microsoft Copilot as a general research tool. The step change came from connecting AI to their data and integrating it into the ticketing system, where technicians already work.
Connecting [ConnectWise] AI Agents to our documentation and ticket history makes PSA much more useful. The search is stronger, the right information shows up faster, and that helps with resolution time, quality of work, and consistency.
Reporting has been an unplanned dividend. Because tickets now carry consistent classification, the team can see recurring issues more clearly. A Tier 1 problem that keeps recurring can be escalated for root cause review rather than handled as another one-off fix.
Better inputs, better outputs, and a product feedback loop
Adoption was a team-wide decision rather than a pilot. It also required a change in habit: closing the resolution field properly on every ticket. The quality of what goes in determines the quality of what comes out. The agents now help close that loop. A technician can drop in notes and have a KB article drafted for the next person who hits the issue.
System Protection Partners treated adoption as a two-way feedback loop. As the team encountered routing scenarios that needed refinement, they shared that feedback with ConnectWise, helping improve the agent experience for themselves and other partners.
We want to keep helping make the product better. If an agent misses the mark, the team reports it, and that feedback helps refine the experience.
Looking ahead
James sees the trajectory as the point. The team is still tuning SOPs, and new agents are arriving. System Protection Partners wants to stay on that path toward more proactive, predictive service.
We're trying to stay on that path because it keeps getting better for us. As new agents arrive and we keep refining SOPs, the service desk gets more efficient.
Key takeaways
- Tier 1 capability lift: CW AI Agents give frontline technicians more Tier 2-level context before they escalate.
- Faster knowledge retrieval: Documentation and prior ticket history are easier to search within the technician workflow.
- Practical time savings: James estimates at least five minutes saved on research-heavy tickets.
- Cleaner ticket intake: CW AI Agents classify, summarize, and route help desk tickets before technicians begin work.
- More consistent dispatch: Routing moved from mostly load-based assignment to SOP-driven, group-aware logic that the team can refine in plain language.
- Better reporting: Consistent type, subtype, company, and summary data give leadership visibility into recurring issues.
- Stronger knowledge loop: Better resolution notes feed future AI output, and agents help draft KB articles for the next technician.
