As a Managed Services provider, Advent One receives 1000’s of client-initiated service tickets a month. Our team is acutely customer focused with a natural tendency for innovation. It’s easy to understand they have a desire to see if we can leverage automation to improve what we do for our clients.
Automation isn’t anything new (particularly for us at Advent One). Most industry peers are looking at how to use AI to improve how client interact with service ticket platforms or to improve the way the ticket platforms communicate with customers.
To really move the needle, we set an ambitious goal of shifting what we do from reactive to preventative.
How? The logical place to start was to look at the data. Which issues were the most recurring across our customer base. We also focused on the service tickets that place the most pressure on our team to investigate, review, respond and remediate.
We looked at metrics over time to understand what was normal or abnormal across the data. For instance, we needed to understand when different companies ran batch processing, which tells us when high levels of activity are normal. We used pattern matching and inference from learnt behaviour to identify where we could make the greatest impact. We saw the sequence of events which led to issues and what could be done to stop them from happening.
We initially identified 8-10 use cases and sites we’ve deployed the automation to. We are already seeing significant benefits after implementation. We’ve reduced human effort required to fix certain issues by 84% (which includes remediation time otherwise it’s 100% as some issues require no human intervention). A simple example is one client running out of storage space. Early action means a system has time to automatically allocate more space before it crashes (a crash can take hours to remediate if it fails).
As always, we have implemented this innovation internally within Advent One first as our testing regime needs to be super rigorous. We are encouraged by the success so far. We’ve created sufficient time savings and capacity within our team to enable us to dedicate a team member to focus full time on this effort.
We’ve started to introduce it to our hosted customers and continue rigorous testing. We observe what the automation recommends or remediates and compare the data to the actions our team actually take to ensure alignment. To be clear, an automated and preventative Managed Services model doesn’t mean ‘eyes off’ – it’s predictive, automated, 24 x 7 with the appropriate amount of human oversight it needs (just less human intervention and oversight). As for our team? They get to focus on deploying even more interesting innovation across the services we offer clients.