Operationalizing AI for Broader Adoption

R. Bhattacharyya

Summary Bullets:

  • A shortage of skilled AI professionals is one of the biggest hurdles to broader AI adoption by enterprises.
  • Salesforce is embedding its solutions with Einstein-powered AI to make the technology more accessible to non-data scientists.

Businesses need artificial intelligence (AI) solutions that can be easily adopted and manipulated by line-of-business users. Although many organizations are eager to adopt AI, they are often held back by a lack of access to data scientists that can curate data and develop AI models that address their specific needs. This skills shortage is one of the top hurdles facing organizations when it comes to operationalizing AI. And it is precisely this challenge that Salesforce is looking to address by embedding its solutions with Einstein-powered AI.

Salesforce is making AI more broadly accessible to non-data scientists, including sales people, account managers, marketing personnel, social media teams, service agents, and customer representatives, by using its Einstein AI platform to power much of its portfolio. The company has outlined four core Einstein capabilities: Discover, Predict, Recommend, and Automate. To discover, users discern patterns in existing data. Once they understand what has happened in the past, they can predict the likelihood of future events and recommend actions based on these insights. The key is to automate the process, so that actions can be triggered as part of a preset workflow instead of requiring manual intervention.

For example, with machine learning, Einstein can use a customer’s existing data to score accounts and flag those at higher risk of attrition (while also identifying the top five factors that have the greatest influence on attrition) and, most importantly, recommend actions. Since the solution is designed to be automated and low-touch, Einstein builds the models, identifies the best model to use, and then uses the data in a company’s CRM solution to score accounts. Once this is complete, it recommends actions.

However, as we all know, the quality of a model’s output depends on the quality of the data used to train it; and the more data used for training, the better the model. Einstein can help novice users with this as well. It will notify users if they don’t have enough data available to build a model. This is critical, since models are developed using only the customer’s data; Einstein doesn’t craft global models using data from other customers’ accounts. While this may limit available data points, it safeguards data privacy and ensures that models reflect the unique circumstances of a specific company’s business experiences.

Salesforce is one of several vendors participating in the movement to drive AI adoption and bring the benefits of the technology to a broader market. Other players, such as IBM and SAP, are also offering solutions designed to target specific business functions. Operationalizing AI requires many components, including intuitive dashboards, solutions that can be manipulated by lines of business, and models that result in actionable insights and trigger workflows, to name a few. Vendors that get the overall combination right will be well positioned to gain market share.

What do you think?

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.