Remote IT management gets a generative AI boost as Atera adds OpenAI Codex

Today Atera announced that it is integrating OpenAI Codex with its RMM platform, to help users automatically generate scripts to help execute processes.

OpenAI Codex is a large language model (LLM) designed to help users with application development. It is a foundational technology that enables the GitHub Copilot service for pair programming.

Remote IT monitoring requires executing a variety of tasks

In the world of remote monitoring and management (RMM) for IT teams, a lot of tasks need to be executed.

Those tasks include system and application management, patch management and resource and storage configuration among others. Today RMM is largely handled by software systems that are increasingly cloud-based. Among the vendors in the space is Israeli firm Atera, which raised $77 million in a Series B round of funding back in 2021 to help advance its efforts. Artificial Intelligence (AI) and machine learning (ML) have long been part of the company’s technology, with algorithms designed to help automate processes and predict potential failures to be remediated.

Connecting different systems and initially setting up some processes however has often required organizations to do some coding to get things to work as desired. That coding effort is about to get a whole lot easier for Atera’s users thanks to AI innovation being integrated from OpenAI.

“What we’ve done with OpenAI is we are releasing a system where instead of writing a script, you just write what you want to do,” Gil Pekelman, CEO and founder of Atera, told VentureBeat. “OpenAI gives you the script and that script with a press of a button is automated.” 

How Atera has been using AI to automate RMM

Pekelman explained that since his company’s inception a decade ago, the basic concept of the Atera platform was to enable users to define IT operations and processes so those operations and processes could be automated.

The processes include daily operations as well as preventative maintenance. The Atera system itself is an IT management platform that combines the technical parts of remote monitoring, patch management and operation automation with the operational side, which includes help desk and ticketing. 

On the monitoring side, Pekelman said that the Atera system collects 60,000 data points per second about the state of an IT environment and its applications. All that data is then used by Atera’s AI algorithms which have been designed to forecast when problems are likely to occur, so they can be automatically remediated before they have an impact.

“Our 11,000 customers run 200 million IT actions every month,” Pekelman said. “Those are actions to install or fix something, and 99.9% of those actions are automated.”

How Atera is using OpenAI Codex to accelerate IT operations

While many common IT processes are already defined in the Atera platform, Pekelman noted that many organizations also have unique and specialized requirements.

To enable the specialized requirements, organizations had to write their own scripts that would then run on the Atera system. Writing those scripts often required time and effort as well as having the right IT skills in place to understand how things work.

With the integration of the OpenAI Codex, the scripting complexity has now been abstracted. Using natural human language, a user just needs to explain what the desired task is and the OpenAI Codex will generate the required script.

The OpenAI Codex itself is a large language model trained on a large corpus of development languages and application logic. Pekelman explained that Atera spent months fine-tuning OpenAI Codex, training the system on Atera’s use case so that it could help solve its users’ challenges.

“We’ve been working on this for quite a while and it’s pretty accurate,” Pekelman said. 

He noted that the generated code might sometimes only provide the user with 90% of what they need, while other times it might be 100%. In any case, he emphasized that the OpenAI Codex code will save users countless hours that otherwise would have been spent writing code manually.

“What it does is it saves hours and hours for an IT person and it also increases the power of the automation of the system that further saves them more hours of work,” he said.