Introducing Comment Summarization: A New Addition to Our Generative AI ✨

Hi folks! 👋

We’re thrilled to announce a new capability added to our Generative AI feature currently available in Wrike Labs: Comment Summarization. This functionality is designed to help you quickly grasp the essential information in long discussions by providing a short overview, listing decisions made, and creating a personalized list of action items 📄

If you already have Generative AI enabled in Wrike Labs, no further action is required; Comment Summary will be available to all users within your account. If Generative AI is not yet enabled, the account owner should first activate the feature. To do this, navigate to Wrike Labs, locate the Generative AI card, and click Enable:

Once the feature is enabled, you will start seeing the AI button in the comment stream header:

Additionally, in conversations with five or more comments, you’ll also see the ✨Summarize comments banner:

Once you click any of these buttons, AI will provide you with an overview of the conversation, decisions made, and personalized action items for you.

Please note that the feature only summarizes text comments in the stream, including:

Regular and external messages
Approvals decision comments (“Added a comment to decision: <text>”)
Proofing comments (only “Added comments” type)

For now, task lifecycle events such as changing status, dates and assignees, are not processed.

We hope that this new feature will enhance your ability to capture the key points from discussions and keep everyone on the same page. We’re looking forward to hearing your thoughts and feedback on this new capability👇

Find more posts tagged with

Comments

Pietro Poli

Very useful for long comments threads

Michelle Hammond

I feel that this could be useful in summarizing long threads as long as commenters are clear in their statements and answers immediately follow the questions they answer (unlike what happens in texts sometimes where the second texter makes a comment or asks a second question before answering the first texter's question, sometimes without re-addressing the first question itself). I don't know if this would confuse the AI or not. I also know that a lot of the people in our facility are not great with spelling, punctuation and grammar! Would things like that throw off the AI? I fear for the accuracy of the AI's summary when people accidentally include typos that appear to contradict what they were trying to say. (For example, instead of saying, "We are now doing this," accidentally typing, "We are not doing this.") Of course, this can cause confusion for the actual humans reading a comment like this as well, but I feel it would cause even more confusion if the AI summarized it into a clearly-worded, concise statement that contained (seeming) authority on the content of the entire conversation. Humans would have a better chance of guessing that it was a typo by reading the context in the actual comment. I have asked my team that if we enable this feature, and if they use Generative AI summarization, that they also read through the comments of the conversation themselves to see if they come to the same conclusion as the AI to gauge if these types of things are happening. It's the only way I feel that the accuracy of Generative AI's summaries could be tested. After we've built some trust through cases like this, then I would put my faith in a feature like this. Also -- it almost goes without saying that including those currently-excluded task lifecycle events would only strengthen the ability of Generative AI to make informed and intelligent statements about the state of a task.

Eric Fiero

As our organization begins using Wrike more this is going to be tremendously helpful. We will be able to quickly summarize comment threads so that it might possibly be included into project reports and end of year reports.

Brooke Langley

Looks like this feature is off, I have seen it pop up. I don't really plan to use it even if it was activated unless I am told I need to.

Gabriel Stangl-Riehle

The new summarize comment tool is going to be extremely helpful when I am meetings and having to explain a thread for others.

Daisy Hibbard

I tested this new feature once and it was really impressive! I'm looking for opportunities in our current workload to test this out more because I can see it being really helpful for our Marketing leaders to quickly get up to speed on a deliverable/project/campaign.

Heather Hernandez

Thanks for sharing, but we aren't allowed to use any AI feature at my company.

Sarah Dungey

This is a really neat implementation of generative AI - creating the TL;DR for longer comment threads on tasks is a cool idea. I look forward to implementing this (eventually).

Ben Thomas

Cool! Excited to try this out

Caitlin Beaty

I am excited to try out to summarize tool!

Angie Cagle

Interesting! I can see that this would be useful.

Elizaveta Bogdanova

Big thanks for your feedback here, folks 🤗

Patrick Jones

Is the information that is generated through the AI cached or saved in a way that is accessible through the API?

Samuel Farrar

Love this and can't wait to try it, as Wrike Admins we spend a surprising amount of time assisting users in interpreting the comments on tasks as others can use difficult to interpret shorthand or they can get a bit baffled by the large number of comments, so anything that helps them and buys us time is very welcome.

Enon Avital

Any way to turn this off as an individual user?

Elizaveta Bogdanova

Hey @Enon Avital, Generative AI is enabled by account owner for the entire account, so it's not possible to turn it off like that. Could you please share why you'd like that?

Enon Avital

@Elizaveta Bogdanova The summary feature is not helpful at all, and pops up on mosts tasks above the comment textarea. It's another bar of bloat in a rapidly shrinking screen real estate, there's already such a small pane for viewing actual task details and it would be nice to permanently hide this unhelpful and prominent bar. It feels like a quick add-on and not something that has been meaningfully integrated, which is a shame and ends up being more interruptive than helpful.

Elizaveta Bogdanova

Thanks a lot for the detailed feedback @Enon Avital, we appreciate it!

Steve Mruskovic

Hi @Elizaveta Bogdanova! I finally got this feature turned on for our team and we've begun to explore it's capabilities.

Today, though, I had a team member use the Summarize Comments feature today and it got stuck in a loop during the Overview, and filled the screen with the same few sentences over and over. Is this a known issue?

Elizaveta Bogdanova

Hi @Steve Mruskovic! I haven't heard of it, so I raised this with our tech support, they will look into this issue and reach out to you separately.

Thanks for letting me know 🙌

Steve Mruskovic

Thanks, @Elizaveta Bogdanova!

Enon Avital

Any updates on being able to disable this on a per-user basis?

Rohan V

Hi @Enon Avital, apologies for the delayed response. Unfortunately, there are no new updates at the moment. We’ll be sure to notify you if anything changes in the future. Thanks for your understanding!