Just Talk to the Database: Boosting Salesforce Productivity With AI

This article was originally published in Thinkers & Makers, a magazine from Akkodis featuring the smartest minds and innovative projects that are driving the future of technology and engineering.

Teaching people how to talk to a database is one of the things Akkodis Salesforce Business Analyst Baptiste Poilblanc and Salesforce developer Eliott Mischler are doing. They utilize Einstein, a suite of AI-based services offered by Salesforce to its users.

The average salesperson cannot extract complex data from CRM without involving a data analyst to write the queries that will produce the requested data.

Ask Einstein Directly

Now, they can ask the Einstein Copilot the exact same question and receive the same data. No more waiting for the data analyst to be available. Instead, Einstein empowers salespeople to access their data directly.

“For instance, you can ask for a closing plan for an opportunity. You give it the name of an account and ask it for help to figure out the next steps", said Mischler. As consultants, he and Poilblanc implement and customize the Salesforce AI tools to suit their clients' specific requirements. That expertise is invaluable, as the Salesforce package can be extended in numerous ways.

Controlling the Ecosystem

Poilblanc stated, “We pride ourselves in having control over the entire ecosystem and trying to make the solutions we build as efficient as possible for the client. We want to give them the best experience possible with these new AI tools.”

The two have connected the Salesforce chatbot feature to an additional knowledge base consisting of small articles with little bits of information. The articles contain information the LLM (Large Language Model) couldn't know without querying the client’s own data. This data could be information about custom discounts based on regions. “

When the chatbot receives a question, it can provide the answer by finding the most relevant knowledge articles, thus being able to give contextualized information to the client. Such a chatbot giving answers based on knowledge articles is not a default feature. We've connected the two over the last few months and successfully built a more efficient and precise chatbot, " said Poilblanc.

The Salesforce Einstein suite is not an LLM on its own. It communicates with an LLM, and you can plug it into any LLM of your liking. On top of that, Einstein can query records, for instance, accounts, leads, or contacts from the Salesforce database.

Users can talk to it, and it will generate a query and produce a set of records that match whatever parameters it was given in natural language. That feature is yet another way to achieve productivity gains via generative AI.

Baptiste Poilblanc, Salesforce Business Analyst, Akkodis

We pride ourselves in having control over the entire ecosystem and trying to make the solutions we build as efficient as possible for the client.

Baptiste Poilblanc, Salesforce Business Analyst, Akkodis

Engaging with Clients

Mischler and Poilblanc focus on the most time-consuming and challenging aspects of the Salesforce ecosystem for their clients. They are now developing AI tools to minimize that trouble and visiting clients to showcase what they have built based on Salesforce Einstein products.

These workshops aim to show clients what they’ve already accomplished and teach them how to customize their solutions to their specific needs. 

“It is all very new, and we haven’t yet discovered all the possibilities”, said Mischler. “Imagine giving a smartphone to a person living in the 1960’s and asking them to do something with it. It is the same thing with AI. The way we approach this is to go out and meet our clients. Together, we can find use cases we wouldn’t have thought of alone.”

Eliott Mischler, Salesforce Developer, Akkodis

Looking forward to New Einstein Features

Mischler and Poilblanc say that the Salesforce Einstein AI suite is only in its very early stages but is already quite powerful. They are now looking forward to seeing even more AI functionality on the Salesforce platform so that they can adapt to their clients’ needs. One such much-anticipated feature is direct voice input. 

“Voice input will be a total game-changer”, Mischler said. “Users will be able to control the CRM with their voice. Let's say a salesperson has just met with one of their clients, and they are on the way home in their car. They can use voice input to tell the CRM via their phone to create a lead with such name and values, schedule a meeting at this date, and more”.

“It’s not a feature that's out yet, but it's something Salesforce has planned. And because it is a huge time saver, we just know that salespeople would do anything to lay their hands on something like that”, he added. 

“That is yet another example of how companies and especially salespeople can achieve significant gains in productivity by turning to Einstein for help. And our most important task is to build the bridges between their needs and the power of these new AI technologies”.