LinkedIn is seeking to make messaging more efficient with its new smart replies feature, which draws on the power of machine learning to offer users predictive text options when in conversation with one another.
LinkedIn senior product manager Arpit Dhariwal explained via a blog post that the social networking site has made its “reply suggestions smarter”.
A short video featured in the post demonstrates smart replies in action, with text prompts automatically appearing in a messaging feed once a message has been received.
For instance, following the question “Hey, how are you?”, the following prompts appear: “Good”, “Fine”, “Great, thanks!”, with the user able to select their preferred option, or instead type out a reply.
“Using machine learning, we’ll show you suggestions that are more contextual and relevant to the conversation you’re having,” Dhariwal advised.
“We’ll suggest at most three responses based on the message you’ve received.”
LinkedIn is set to further personalise the feature by increasingly tailoring suggested replies to the conversation a user is having. Dhariwal said users “may soon see a response like ‘Thanks, Joe!’ versus the more simplified ‘Thanks’”.
Users can choose whether or not to use the feature by going to their settings.
The feature will first be rolled out in English globally via LinkedIn’s mobile app and on desktop, with Dhariwal saying other languages will be coming in the future.
Never miss a story: sign up to SmartCompany’s free daily newsletter and find our best stories on Twitter, Facebook, LinkedIn and Instagram.
Martin Kovacs is a journalist with experience covering the IT, consumer electronics, retail, finance and energy sectors.
