25 years after Salesforce was founded, the company is focusing on AI. What Salesforce does and how it does it has evolved considerably over the past few decades. However, the underlying core values ​​remain the same.

The architecture of the Nekkerhallen in Mechelen is not really suitable for the exciting drum band. For ten minutes the drums echo through the cozy hall at a volume that the trumpeter present has difficulty drowning out. It is unforgettable and by the time Lien Ceulemans, Country Leader for Salesforce in Belgium and Luxembourg, steps onto the podium, it is already clear to those present that the World Tour Essentials is not a standard IT event.

Birthday

After all, Salesforce has been around for 25 years this year. The CRM specialist has been present in Belgium since 2010, and Ceulemans has been at the helm there for three years. A lot has changed in all these years. Salesforce is listed on the New York Stock Exchange as “CRM”, but today it offers its customers much more than just a Customer relationship managementSuite. Data and the insights that can be gained from it, often with AI, are of central importance.

“You have to build a relationship with your customer Customer 360 “This is necessary,” says Deputy Country Chairman Gianni Cooreman. He refers to Salesforce’s strategy of collecting all data about a customer in one place, regardless of whether it comes from sales, the help desk, the web shop or elsewhere.

“We have been innovating for 25 years to achieve this,” he continues. “With the data cloud, we now have a scalable platform into which you can bring external data. We have finally achieved the Holy Grail, a only source of truth In the hands.”

Just in time for AI

The timing couldn’t be better: A central source of correct data is the foundation of AI projects and AI is currently an important pillar for Salesforce. “AI is no longer a buzzword,” says Ceulemans. “It is the core of our platform, for SMEs and large companies.”

AI is at the heart of our platform for SMEs and large enterprises

Lien Ceulemans, Country Manager Salesforce Belgium and Luxembourg

Cooreman sees four waves of AI development. “The first wave was predictive analytics. We introduced this capability with Einstein and it is still very important. The second wave came with ChatGPT. The third wave is just around the corner. AI agents will talk to each other. And then one day the fourth wave will come, with general artificial intelligence. It will happen, the only question is when.”

Today we are in the middle of the second wave and that also brings challenges. Because how do you build good, reliable and ethical AI applications as a company? “You shouldn’t use public data,” says Cooreman, “It won’t help you improve sales and marketing in your company.”

Optimistic

“You need to connect models to your data from all systems, from CEM and ERP to website and product data to the mainframe. That’s how you make it relevant.” Even then, Cooreman still sees challenges, such as hallucinations caused by generative AI models or distortions in the training data. “AI models are convincing liars.”

However, Salesforce has great confidence in the future of AI. “If we get this right, and we will get it right, the benefits to society and the economy will be enormous,” it says in a sensational video in which terms such as “certAInzig and extraordinaryAIshoot over the picture.

AI with core values

Ceulemans explains what exactly that means and why Salesforce is positioned to lead a responsible AI revolution. “I truly believe we are part of a broader ecosystem that contributes to the economy and the world,” she says. “The values ​​we convey are real. That’s why I work here. We are more than just a technology company.”

We briefly summarize these values ​​for the uninitiated: They revolve around trust, customer success, innovation, equality and sustainability. The five values ​​are particularly relevant when developing an AI strategy. Innovation alone is not enough, but must be tailored to the customer (and their data), trust-building and unbiased, and always with an eye on sustainability.

Trust layer

Trust is a big part of the event. Robb Hufschmitt, Senior Solution Engineer, explains what Salesforce actually does. “We don’t train LLMs,” he says. “We want to train AI with business data, but an LLM doesn’t have security built in. If you use data for training, anyone with access to the LLM can extract it, and we don’t want that. In addition, an LLM always tries to give an answer even if there isn’t one,” he says. “That’s how you end up with hallucinations or a toxic response.”

The Einstein Trust Layer solves this problem in a way that combines 25 years of Salesforce innovation and values: The layer makes a prompt to an LLM more intelligent and enriches it with relevant company data from the data cloud in which the only source of truth lives from internal and external data. The trust layer ensures that an LLM of its choice receives the right prompt and the right data and returns reliable answers. “Data, plus CRM, plus AI, plus trust: that’s the be-all and end-all today,” summarizes Hufschmitt.