Photobank - Fotolia

Three ways Salesforce Einstein can enhance workplace analytics

Smart cloud analytics platforms such as Salesforce Einstein play well as flashy demos, but they can have a substantive impact on day-to-day work in the enterprise.

Last fall was heavily populated with impressive smart cloud launches, from Microsoft Azure to IBM Watson Analytics and Salesforce Einstein. Demos presented at the major annual tech conferences were flashy and promising.

Microsoft has chosen to bank on its build-your-own philosophy, an approach that did well for SQL Server. IBM Watson, though presented as a workplace analytics slot machine of sorts, is nonetheless wedded to an important array of existing products and functionality. Salesforce Einstein, presented at Dreamforce 2016 in San Francisco, was tentative -- bombastic, but unfinished, though bursting with potential.

But Einstein is crippled by the limitations of its platform. Salesforce, primarily a sales and marketing environment, lacks IBM's processing breadth and Microsoft's functional depth. Number crunching fireworks aside, can Einstein make a difference in day-to-day business operations and processes? A review of its architecture, its approach to social media integration and its suite of new APIs seems to indicate it can.

New power in old apps

As its app base expands, Salesforce can now offer artificial intelligence (AI) integration to emerging apps that can positively affect a number of routine business processes. The functionality is available throughout the platform architecture, in full-code apps, low-code apps and point-and-click apps -- the goal being AI-first apps.

AI-augmented apps on the Salesforce platform could potentially bolster daily performance in two prominent ways.

  • By reducing the time required to research and create new content. This is achieved via APIs that leverage Salesforce Einstein functionality in order to classify unstructured input and to passively apply image recognition and product recognition. Manual lookups become unnecessary. This is the kind of passive functionality that rapidly becomes so convenient that it's taken for granted.
  • By embedding Einstein's prediction functionality, with recommendation capacity built in. This relieves users from the task of active data validation, except where absolutely necessary; best-of-class data and content are presented and pre-evaluated. This predictive functionality also works on social media streams, scoring customer sentiment as it comes in, and could even be used to trigger workflow -- another day-to-day time savings.

Salesforce Einstein and the Community Cloud

Predictive sentiment isn't the only AI application affecting Salesforce's social media functionality. The Community Cloud, already an interactive environment for socially assisted content discovery, could leverage Einstein AI for faster access to the best information.

AI-driven community engagement, a long sought-after holy grail for platforms like Microsoft's SharePoint, can include relevance ratings of trending pages and posts in the Community Cloud, hastening real-time access to not only content, but also to human experts and persons presently engaged in important topics -- another huge time-saver in day-to-day work. The same mechanism can facilitate discovery of new community members, groups and resources based on not only expertise, but also engagement.

Looking ahead with Einstein

Finally, Salesforce Einstein can enhance operations, such as big data as software as a service and machine learning for ad hoc data analysis, with its predictive services, its most generic workplace analytics feature.

The advantage here is that the ad hoc data analysis can be configured for platform events, transactions in Salesforce that can be tagged as workflow triggers in existing data streams. They're calling this event-driven intelligence, and it's another consolidation of functionality that stands to save lots of time in daily operations. It's one thing when transactional processing triggers new events; it's another thing when workplace analytics based on those transactions are doing the triggering.

This functionality should also be portable to the internet of things (IoT) cloud, making it possible to perform real-time quality evaluation on inbound IoT data -- removing yet another manual task from the workday. 

Next Steps

Salesforce Einstein vs. Microsoft AI: A discussion

How Salesforce IoT Cloud and Einstein work together

An expert looks at Salesforce Einstein and the industry

Discover how AI improves Salesforce analytics

Dig Deeper on AI for sales