As managed service software vendors look to harness the power of AI and machine learning within their platforms, MSPs can expect to see a gradual, yet significant, emphasis on automation.
Automation is already deeply tied to the managed service business model, which requires continual improvements to service delivery to grow recurring revenue. MSPs with limited staff can use specialized tools, such as remote monitoring and management (RMM) software, to support a large volume of customers. However, while RMMs and other managed service software can efficiently drive many day-to-day MSP operational processes, a portion of technicians' time and energy remains tied to labor-intensive chores.
"Software for MSPs has always existed, really, for one reason: to automate manual tasks," said Mike Puglia, chief marketing officer at Kaseya Corp., an IT infrastructure management software vendor that targets MSPs.
In the future, MSPs could, potentially, assign many of those manual tasks to an AI system, freeing up human technicians to create more valuable offerings for customers.
How MSP software embraces AI
Kaseya, among other managed service software vendors, has started to explore how AI and machine learning models can improve its tools, with an initial focus on cybersecurity. In August 2020, Kaseya acquired Graphus, a cloud-based email security provider. Graphus uses machine learning to automate the prevention of email-based attacks.
MSP software vendor ConnectWise also applies AI and machine learning to cybersecurity within its products, mainly to detect anomalous user behavior. "We have built up some models, and we are utilizing the ingestion of different data points to look for [red flags]," said ConnectWise Chief Product Officer Jeff Bishop. ConnectWise's current AI-based security efforts focus on the company's RMM and remote-control technologies.
ConnectWise has begun testing how AI can enhance help desk ticketing and service delivery for MSPs, Bishop said. For example, the company is exploring how AI could automatically route help desk tickets to the right person or team. ConnectWise has also experimented with sentiment analysis to prioritize help desk tickets, escalating tickets when the system detects anger in the ticket filer's tone.
"These are things that [are mostly in an] alpha or beta state," Bishop noted.
To grow their businesses, MSPs must tackle increasing levels of operational complexity. Learn how different automation tools and approaches can reduce time-consuming tasks.
ConnectWise partners with several third-party vendors of chatbot or service delivery tools with AI built in. MSPs can integrate those products into the ConnectWise software. "We have done our best to make sure that, through our APIs and integrations, we support [that technology]," Bishop said.
Pulseway, which provides RMM tools for MSPs, is also taking a measured approach to incorporate AI into its software. "We are going in for the most tangible [applications] of AI technologies inside our product," said Pulseway Founder and CEO Marius Mihalec.
As part of that initiative, Pulseway has built a machine learning-based notification and alert system into its RMM. While still in its early stages, the system aims to analyze MSPs' customer IT environments to predict technical problems and alert the MSPs' techs when they need to respond. "The accuracy is not where it should be, but, in most cases, we do provide a feedback mechanism" where an MSP can indicate whether a notification or alert is valuable, Mihalec said. This feedback helps train the machine learning model.
Proceed with caution
Despite AI-based automation's potential, MSP software vendors recognize there is a long way to go before the technology matures. Until then, it requires a tight leash.
ConnectWise's Bishop said AI has already demonstrated bad decision-making in several cases. "We are allowing [AI] and machine learning models to make decisions, and it is a little bit scary -- or it can be -- if you don't have full control over it," Bishop said. "We want to be very careful and cautious of how we roll this out."
Puglia and Mihalec shared Bishop's concern.
"One of the biggest fears of people in IT … is that AI can do a lot of things that could wreck a lot of [things] by mistake," Puglia said. "If you don't have any kind of guardrails," the AI system could be more foe than friend.
"Obviously, it is a bit scary," Mihalec said. "This is a factor we take into account." He noted that Pulseway's notification system will automatically suggest, rather than execute, an action, because each customer IT environment is different and MSPs can handle each issue in various ways.
That said, Kaseya, ConnectWise and Pulseway all view AI as increasingly playing a role of an assistant in MSP operations. For example, AI systems could help technicians identify patterns and suggest ways to resolve IT issues, potentially boosting technicians' productivity.
"I think that is where we are going for at least the next 24 months," Puglia said.
Future implications of AI-based automation
Down the road, as AI technology matures, new automation capabilities could free up MSPs to focus more on their businesses, the MSP software executives said.
According to Puglia, MSPs are typically stuck in the operational weeds, which hinders them from growing their businesses. AI-driven automation could change that. "Every MSP I have talked to [says] they could do so much more if they could get out of a lot of the lower-level stuff, " he said. Once liberated, MSPs would be able to do more consultative-type work and, ultimately, be more profitable.
Bishop agreed MSPs will eventually see considerable benefits from AI in managed service software. "We see [AI] playing a major role in helping [MSPs] provide better customer success, as well as just be more efficient in the business to help them improve margins, be more profitable and really focus on their end clients," he said.
Mihalec said he believes future AI advancements could fundamentally change what MSPs do.
With an AI-powered RMM, for example, MSPs could configure the system's level of AI automation according to each customer's environment, then simply monitor the AI's actions against the applied policies. As a result, the MSP would become more like an "observer," ensuring the RMM system complies with a customer's service-level agreement. Additionally, an AI-powered RMM would enable MSPs to provide data-driven consulting services, using the data the system collects from the customer's IT environment.
"That is where [RMM] technology is going in our view," Mihalec said.