Salesforce uses AI to anticipate your sentiments and intentions

Salesforce has upped its efforts in the artificial intelligence world, introducing three new features for developers in its continued quest to ‘democratise’ the technology.

Previous IDC estimates have certainly put forward a strong case for AI and it would appear Salesforce is doing everything possible to capitalize on the trend. In a recent report, IDC has claimed that 80% of apps will contain some element of AI by 2020, but the challenge still remains in that it is a very complicated world to get to grips with.

The new updates will allow Salesforce developers to build AI on-top of any CRM apps which they build through the platform, essentially removing some of the complications. The objective here is to build ‘personalized, predictive customer experiences’, as well as create models based on deep-learning to continuously improve performance. It’s a bold quest.

The first update, Einstein Sentiment, will allow developers to create tools to understand the context and emotion of text. Whether this is an email or a social media post, or even a review on a message board, the tone of the message can be identified as being positive, neutral or negative, which in theory, should dictate some sort of response.

The second, Einstein Intent, will allow developers to train a model to classify the underlying intent of customer inquiries. Combined with the first update, more accurate automated responses can be created, allowing for customers to more effectively use self-service portals, or even pave the way to more creative and responsive bots.

The final update, Einstein Object Detection, does exactly what it says on the tin. It can look at an image and recognise location, size and quantities of different objects in said image. It sounds simple, but in terms of inventory management or similar applications, it could be a very useful proposition.

The tools themselves will allow developers to create a more customized experience for their organizations. The armoury of APIs is starting to look very strong at Salesforce, but the team are also enabling the integration of historical data to make these predictive and reactive models more accurate for the user.

Salesforce has long been the leader in the CRM game, and while there are challengers who wish to disrupt this position (Oracle being the most vocal), there are few which are putting the same effort into intelligence. The others are playing catch-up, but Salesforce is changing the rules. Intelligence is looking like the new battle ground for the connected economy, and Salesforce is certainly putting itself in a strong position.