Slack adds AI-powered search and summary to the platform


As a business communications platform, Soft has become a de facto repository for storing institutional knowledge, but it is difficult to access this information with conventional search tools. Today, Slack introduced a few new features designed to make this information more accessible, including a new AI-powered search tool and the ability to summarize information in channels.

Noah Weiss, director of product at Slack, says the platform naturally gathers company information in an informal, unstructured way. The problem is finding a way to bring out this hidden treasure trove of knowledge. “The great thing about all of this is that this new wave of generative AI capabilities now allows us to extract a whole new set of meaning and intelligence from all of these analytics that have been created for years. [on our platform]”Weiss told TechCrunch.

Last May Slack announced that he was integrating generative AI into the platform during the Salesforce World Tour in New York. Rather, it was a widespread call to action with the creation of SlackGPT, its own version of generative AI designed specifically for content on the Slack platform.

Today’s announcement aims to implement this more specifically. Weiss says being able to summarize a channel’s content helps employees catch up after a leave of absence, or simply avoid having to read a long thread to get the gist of the conversation. With channel summaries, you can request a summary and Slack’s AI model generates a summary of all the topics covered along with references to show how the model created each part of the summary, which Weiss said was an essential part of the design of this feature. .

“You can explore any area and we show you all the detailed context. So we were really thinking about transparency, building trust, making sure we show our work and giving people the opportunity to delve deeper to find out more if they want to,” he said. declared.

Gif showing how the Slack Summarize AI feature works.

Image credits: Soft

The company also allows users to ask questions in a natural way, just like with ChatGPT, but it uses Slack content instead of more generalized Internet content, so a user can ask a question like “What is is Project Gizmo? SlackGPT then provides a response, again with sourcing, to allow users to see where the response came from and whether they can trust it.

Slack AI search asking what the project gadget is and giving an answer derived from the Slack answer archive.

Image credits: Soft

Each response includes a quality check, where users can tell whether the response was good, bad, or neutral, so the model can learn about the quality of the responses and system engineers can see how the model is performing.

He wouldn’t go into detail about the underlying model, saying only that it was a mixture of large language models. “What we found is that they all work in different ways and with different speed and quality characteristics. We spent a lot of time fine-tuning the models for the data we actually have in Slack, and we also did a lot of work on rapid engineering.

Slack AI with Search and Summary is an add-on product for enterprise plans, which means it will cost additional money on top of the normal licensing cost. Slack did not provide details on costs, but it is available today in the US and UK, in English only for now, but will be available in other languages ​​in the near future, according to the society.


Leave a Comment

Your email address will not be published. Required fields are marked *