An AI-powered social marketing tool called PI has launched.
The new AI tool was developed by San Francisco-based Post Intelligence – founded by ex-Google executives Bindu Reddy and Arvind Sundararajan.
The AI tool creates a “deep neural network” for each user, analyzing their posts to determine when most followers read them, it then builds custom models of engagement prediction.
“Whenever your followers are the most engaged, an optimized time slot is created for your account. PI then provides up-to-the-minute suggestions of content to fill that slot,” Post Intelligence said in an official blog post announcing the new agent.
“Using its powerful AI engine, PI knows what kinds of topics and micro-topics your followers are most interested in, and puts content that checks all the boxes at the top of your feed.”
Deep-learning allows the AI tool’s skills to get better with time, “so the more you use it, the better it will get,” the developers say.
Users can post on multiple pages on different platforms through Pi and schedule when the posts will run.
Currently there is support for Facebook, Twitter and Pinterest. The team is reportedly working on getting the AI agent to work for Instagram and Snapchat as well.
Post Intelligence provided the following example of how PI functions.
If, for example, a user asks the AI tool to tweet about how unhappy they are about a new healthcare bill PI can help construct the tweet on their behalf.
“Hi Pi, Post a tweet about how disappointed I am with the new healthcare bill and millions losing insurance”
PI would understand the meaning behind the sentence and collect real-time find data about the healthcare bill to create a tweet with a link to an article about healthcare. It could post something like:
“Totally bummed that 20+ million will be losing health care under the new GOP plan http://healthcarelink”
“We believe the future of publishing will be curating your posts the same way you curate your feed — adding and subtracting from your mix and focusing on what brings the most joy to your followers. That means a variety of different tools, support for new social networks as they emerge and better data about what’s worked well in the past,” the PI team said.