Pia Pulls Humans a Little More Out of the Loop
The company’s latest step on a long, carefully-paced march toward autonomous AI for service desks is a customer-facing chat tool capable of closing most Level 1 tickets on its own.
Patience has long been Pia’s hallmark. One of the earliest names in service desk automation software for MSPs, the vendor has carefully held autonomous ticket closing powers in reserve to avoid pushing MSPs outside their gradually expanding comfort zone.
Too often in AI, explains David Schwartz, the company’s CEO, “it feels like you just set out to go climb to the peak of Everest and the reality is you don’t have a map, you don’t have a compass, you don’t have your stops along the way. It’s just get to the top. We’ve been very purposeful in saying that doesn’t work.”
One step at a time does, he continues. The company’s latest step, announced during pre-day activities at Kaseya’s recent Connect event, aims to make resolving a service issue less painful for both the client experiencing it and the MSP addressing it. Called Pia Chat, the new feature lets end users report a problem via natural language chat from directly within Microsoft Teams.
“You don’t have to leave that tool that you’re used to using every day,” Schwartz says.
Neither do help desk technicians on the receiving end of a ticket, who can close the ticket based on detailed contextual information collected automatically by Pia from inside their PSA system—if they need to touch the ticket at all. The system can resolve simple issues like password resets on its own, sometimes within seconds.
“Most Level 1 requests can absolutely be done without any human in the loop,” Schwartz (pictured) says.
That step forward in Pia’s steady march toward fully automated ticket resolution comes in response to the growing acceptance of autonomous AI among MSPs signaled, among other things, by their increased use of the company’s AI AutoStart feature.
“It does a lot of the troubleshooting on endpoints and other work that a tech would normally have to do,” Schwartz says. “We are now seeing more partners leverage it.”
That increased willingness to let AI do unsupervised help desk work is new since the last time Schwartz spoke with Channelholic at Kaseya’s DattoCon event last October. “There’s been a noticeable change,” he says. “There’s an understanding that we want to move into this agentic world, but we want to know that it’s guardrailed.”
Pia is similarly cautious about autonomous AI. “If it’s used right, it’s a force multiplier. If it’s used wrong, it’s just fast-tracking to a bad result,” Schwartz says.
Pia Chat, he continues, is an example of AI done right, as well as an illustration of Pia’s belief that “practical AI” is better than AI features that look impressive but have little impact on MSP workflows.
“It’s not AI for the sake of being AI,” Schwartz says of Chat. It’s AI designed to provide a specific outcome with a measurable impact on both technician productivity and end user satisfaction. More such features that reach beyond the basic blocking-and-tackling of closing tickets to help MSPs forge stronger client relationships are currently on the roadmap.