For many visionaries and technology firms, artificial intelligence (AI) is the future. But, today at the TrailheaDX developer conference in San Francisco, Salesforce brought it into the “now,” showing off its AI-first Salesforce Customer Success Platform.


With advances in machine learning, deep learning and other AI technologies, the technology industry is moving from systems-of-record and systems-of-engagement, to systems-of-intelligence to help our customers better connect with their customers.


Salesforce is leading this AI-first customer approach, and on stage at TrailheaDX we walked through a scenario of a buyer in the market for a house on a popular real estate site (and a Salesforce customer). Today, services are connected not only to buyers, sellers and agents, but also to Salesforce and Internet of Things (IoT) data -- making the customer experience even more social, mobile, connected and intelligent than ever before.


For this demo, we built an app on Heroku that connects to Facebook Messenger, with AI-powered bots from Salesforce delivering real-time conversations with the real estate service. A message from a potential buyer on Facebook Messenger is pared by a bot, which then makes a query to Salesforce for any relevant information, such recent homes that were viewed and favorited.


In this example, the buyer specifically uses Facebook Messenger to ask the real estate service for recommended listings based her profile. The real estate service shows the relevant listings, but the buyer’s profile on the site doesn’t represent her real interests. The bot then asks if the buyer has a Pinterest board, which she does and is able to share with one click. The bot then uses the Pinterest API to go through the images and links them to MetaMind, Salesforce’s deep-learning platform, accessing a pre-trained classifier that categorizes the style of houses as Colonial, Contemporary, Victorian, Tudor or Greek Revival.


The data on which type of architecture the buyer prefers -- based on her Pinterest images -- is passed onto Salesforce. Then, the relevant properties are presented to the buyer. Another part of the bot schedules private home tours, creating a record in Salesforce that connects the agent to the buyer. All of this is done in the blink of an eye, in the background, seamless to the end-user.


What about when the buyer visits the house in person? Lockitron, an IoT solution, allows the real-estate agent to unlock the door from his mobile phone. And when the buyer wants to walk around the house without the agent tagging along, Amazon’s voice-controlled Echo speaker and digital personal assistant Alexa are on hand to help. For example, the buyer can ask Alexa about nearby amenities that match her interests, such as local restaurants or grocery stores, and receive a Walk Score detailing the “walkability” of any address in the neighborhood. And the raw text of conversations with Alexa is passed onto Salesforce, so if the buyer asks Alexa about the flexibility of the seller on the price, Alexa alerts the broker of a potential offer.

In this scenario, Salesforce’s Thunder IoT cloud is also continuously looking at events from multiple inputs, listening to state changes and responding appropriately. For example, if the number of “likes” on a property crosses a threshold of 100, Thunder triggers the scheduling of an open house and sends a notification to all people who “liked” the property.

This is just one example of how the customer experience will be transformed by an AI. Not in the future -- but today -- as these business scenarios become a reality on the Salesforce Customer Success Platform.