
LogicMonitor is partnering with OpenAI to apply generative artificial intelligence (AI) to IT service management (ITSM).
As a provider of a software-as-a-service (SaaS) platform that is widely used to monitor IT environments, LogicMonitor is able to collect the types of data that would be needed to train generative AI models that could be employed to automate a range of ITSM tasks. That data has already been used to create Edin AI, an AI agent that LogicMonitor created to streamline the number of alerts generated by its platform.
Karthik SJ, general manager for AI at LogicMonitor, said the company is making use of the open source Langchain framework to train multiple types of large language models (LLMs), including an o1 model from OpenAI that has advanced reasoning capabilities. The overall goal is to apply generative AI to ITSM tasks in a way that will culminate in the development of AI agents that ITSM teams will be able to orchestrate.
The overall goal is to not just automate tedious tasks but also elevate the ability of IT teams to manage environments at greater levels of scale, said SJ.
Itโs not clear exactly what impact AI will have on ITSM, but there will undoubtedly be disruption. On the one hand, many tasks currently performed by IT administrators will be automated but, in truth, not many IT administrators especially enjoy those aspects of the jobs.
Advances in AI should enable those IT administrators to manage systems and applications at much higher levels of scale, said SJ. That’s critical because there is still a general shortage of IT expertise that limits the number of applications any organization can deploy, he added.
In fact, AI should enable smaller organizations in the age of digital transformation to compete more effectively against larger organizations that are better able to employ small armies of IT personnel, noted SJ.
Itโs not clear how long it will take for this AI future of ITSM to arrive. Rather than being a seismic event, itโs more likely that multiple AI agents capable of automating the management of various tasks will be continuously added to various platforms. The challenge then will be finding ways to orchestrate the management of those AI agents to ensure an end-to-end task is completed. Given the probabilistic nature of the LLMs used to train these AI agents, there will be a need for humans to verify that tasks are being completed in a way that meets expectations. The core issue is that many IT processes are deterministic, so they often need to be completed the same way each time, which is not necessarily something an LLM is designed to provide.
Like it or not, the proverbial AI genie is not going back into the bottle. ITSM teams should start determining which tasks being performed today might lend themselves to being automated using AI agents tomorrow. Just as importantly, however, there may be tasks that are not being done at all, simply because it wasnโt feasible for an IT team to handle. In the age of AI, there is now an opportunity to revisit the art of what is possible that only comes along but once in a career.