InvGate’s AI Hub generates knowledge articles from IT incidents

Join leaders in Boston on March 27 for an exclusive night of networking, insights, and conversation. Request an invite here.


Once a company reaches a certain size or level of complexity — particularly in regards to the processes that its employees and contractors are assigned to go through — it may be prudent for the firm to start maintaining a knowledge base of articles explaining how to do things and resolve common issues. 

Whether it’s an internal website, corporate wiki, or even just a set of documents in the cloud or on paper, such a knowledge base can help employees at companies find out instructions on how to do common tasks or resolve issues, such as onboarding, setting up a new machine, downloading appropriate software, requesting and booking time off, or even troubleshooting their workplace apps and files. The knowledge database can also help developers at a company determine how to write and debug their software, and ideally, offload some of the work of poor overextended IT support teams.

Yet the process of creating such knowledge base articles has traditionally been a painstaking and surprisingly antiquated one — tapping writers internally or externally with some degree of experience or expertise in translating technical subject matter into digestible directions. At least, it has been until InvGate came along.

The 16-year-old Buenos Aires-based company recently announced the launch of its new AI Hub, a tool powered by leading large language modes (LLMs) (chosen with user input) that automatically turns incident response conversations between an enterprise’s IT support staff and employees or customers into knowledge base entries.

VB Event

The AI Impact Tour – Boston

We’re excited for the next stop on the AI Impact Tour in Boston on March 27th. This exclusive, invite-only event, in partnership with Microsoft, will feature discussions on best practices for data integrity in 2024 and beyond. Space is limited, so request an invite today.

Request an invite

“We allow our agents to select from which incidents they want to create knowledge, and we draft a knowledge base article to be published automatically,” said Ariel Gesto, Invgate’s founder and CEO, who previously worked in IT operations, in an interview with VentureBeat. “But then, we apply AI summarization to allow the knowledge able to be consumed in a conversational interface.”

In other words, IT team customers of InvGate who choose to access the new AI Hub (it’s available as an included update in current pricing) can select which incident response conversations they wish to turn into knowledge base articles, run them through the InvGate AI Hub tool, and out comes not only a new article, but a conversational chatbot version of it that you can ask questions and receive responses about that specific type of issue.

A Hub controlled by users

In order to access the AI Hub, customers already have to be using InvGate’s older IT Service, Enterprise, or Asset Management platforms and Service Desk software. The offerings start at $300 per year per support agent.

“In the majority of cases, we are consuming our own sources of information related to incidents solved in our platform,” Gesto said.

This approach not only ensures the AI Hub has access to the most relevant information for crafting knowledge base articles for a customer, but also that the information remains secure and encrypted in its SOC 2 compliant systems, a leading third-party security standard.

Furthermore, customers have control over what information they share with the AI Hub and all of InvGate’s offerings, according to Gesto.

“For example, if you’re doing something related to HR [Human Resources] and don’t want that information fed into a large language model, you can select not to include it and filter the information you want to include from what you don’t,” he told VentureBeat.

Moreover, InvGate uses both Microsoft’s Azure OpenAI cloud service and local versions of open source models such as Meta’s Llama 2 running on its own servers to provide the LLM backend powering the automatic knowledge base generation of the AI Hub.

Customers wary of having their data sent to OpenAI models — though the company explicitly says it doesn’t train on content submitted by ChatGPT Enterprise and Teams users — can elect to use the open source offerings instead. In fact, this is central to InvGate’s approach — building a flexible system that can accommodate multiple models, even future ones yet to be released.

“We’ve gone ahead and built our own layer service that we use essentially so that, depending on what type of ask it is and what task needs to performed, we can use any number of AI models,” said Nacho Harriague, director of product marketing at InvGate, to VentureBeat.

The AI Hub also allows end users to control whether they want to read the knowledge articles it generates on their own, or engage with them in the conversational interface, giving them the option that best suits their needs.

A strong track record and impressive roster of customers

InvGate’s track record of providing reliable and useful IT asset management and customer response software has already won it a sizaable base of users, including such notable organizations as NASA, McDonald’s, PwC, and Toyota. Altogether, InvGate claims to “support over 1,000,000 employees across 50 countries.”

Many of these companies switched to InvGate from other legacy IT support management systems, according to Gesto.

InvGate is an “omnichannel” movement that allows its customers to create support tickets from a number of sources, from email to Microsoft Teams to its own API. Meanwhile, InvGate provides a Virtual Agent that plugs into Microsoft Teams and can access the knowledge base and provide answers in a conversational format — though this is available through the InvGate Service Desk, not the Microsoft Teams AppSource store.

With the launch of AI Hub, the company plans to further build on its success and offer new Gen AI capabilities to increase its customers’ efficiency and help IT teams solve problems faster, with less time spent searching and retrieving information.

Originally appeared on: TheSpuzz

Scoophot
Logo