Salesforce reckons it’s the end of the DIY AI era – and global CMO Ariel Kelman is tasked with addressing what his CEO, Marc Benioff said last week is Salesforce's biggest marketing challenge: convincing global markets to think less about Open AI, Microsoft copilots and other generative AI companies that require businesses to custom-bake the tech into their organisations to make it work – and more about the deployment of low code, no code, autonomous AI agents that can be built and tested and live within weeks, if not days.

The difference is that Salesforce is pointing these agents directly at existing customer systems and data, rather than brands spending “literally tens of millions of dollars with cloud providers to train these models” from scratch.

“There are lots of use cases where you do need to train and fine-tune your models. But absolutely not sales, service, marketing and commerce – the models are smart enough that they can go and grab information,” says Kelman. “It can just scale the work that our customers have already done.”

It’s working for the likes of Saks, Gucci and Wiley – and some local firms like Fisher & Paykel and Queensland University of Technology are now likewise plugged into what Benioff reckons is “AI’s third wave”.

Kelman says AI agents “blur the lines” between sales, service and marketing functionality – and coming next is a variant for sales lead development, where the agent will develop the leads until they are warm enough for a human to take over.