By: Henry Kaiser
Einstein and Watson Collaborate for a More Powerful CRM
Two of the world’s largest providers of customer relationship management (CRM) solutions are joining forces to expand the scope of their artificial intelligence systems. IBM applied a significant portion of their resources to develop Watson, a cognitive computing technology capable of analyzing almost any kind of data, from text and images to audio and video information. Just last year Salesforce released their new AI-powered CRM add-on called Einstein, designed to provide deeper insights for sales, marketing, and customer support teams.
Now Einstein and Watson are being combined to produce the best recommendations and predictions for client companies based on advanced machine learning algorithms. By integrating Watson’s application program interfaces (APIs) with the cloud-based CRM software from Salesforce, the combined predictive analytics can provide smarter, faster decisions.
The Watson APIs are designed to deliver predictive insights from unstructured data, inside or outside an enterprise, data that can include everything from weather patterns, supplied by the IBM Weather API, to shopping trends and ongoing relevant market activities. Einstein is tailored to solicit predictive insights from customer data inside the client’s Salesforce CRM. By combining the two, a company like REI, for example, could know when a series of rainstorms is predicted near any of their outlets and send personalized marketing emails to customers living in the path of the storm and likely to need new waterproof gear.
The IBM Watson and Salesforce Einstein integration should be available in the second half of this year, and Ginni Rometty, chairman, president, and CEO of IBM has predicted that, “Within a few years, every major decision – personal or business – will be made with the help of AI and cognitive technologies.”
Smart Shoe App
AI-powered applications are being employed by an expanding array of retail businesses to better engage with their customers. One AI-powered app called Shoegazer is designed to enable users to identify shoes that they spot on the street and want to buy themselves. Just like Shazam recognizes songs by listening to a few bars of the tune, Shoegazer will analyze any photo you take of the shoes you’re interested in and tell you the brand.
AI developer Happy Finish announced that Shoegazer is currently at the proof-of-concept stage, with the goal of matching any photograph taken by the customer with specific brands and models of the shoes in real time. The app employs image recognition and transfer learning to identify the make and model of each shoe, and is just one example of the kind of applications that could leverage AI and machine learning to sell a variety of products. The app can currently identify popular trainers with an accuracy approaching 95 percent.
A Cognitive Platform for the Private Cloud
Machine Learning is often considered a sub-set of artificial intelligence that uses computer algorithms to autonomously learn from data and improve those same algorithms automatically. Machine learning is one of the key technologies inside autonomously driven cars, facial recognition systems, and biometric identification. Last month IBM announced IBM Machine Learning, as the first cognitive platform for continuously creating, training, and deploying a high volume of analytic models in the private cloud.
IBM extracted the core machine learning technology from IBM Watson and will initially make it available through the IBM z Systems mainframes, which provide the operational core for most of the world’s banks, retailers, and governments, covering the billions of transactions processed by those entities daily.
The Three Stages of Machine Learning
The editors of destinationCRM.com recently posted an article suggesting that one of the best ways for a company to leverage a CRM investment and improve ROI is to employ machine learning as an intelligent layer on top of their existing CRM systems. Machine learning can draw insights from all of a company’s data to uncover every customer’s full story and determine the best way to maximize their sales.
The article concludes that this can be accomplished thanks to the power of what the they call The Three Stages of Machine Learning:
• Analyzing the past to understand what actions have led to the best outcomes
• Interpreting each new customer interaction and making recommendations for the next best action to produce a successful outcome
• Continually updating the machine learning application based on the most recent set of sales outcomes in order to stay relevant in real time