More than 10,000 people flocked to Salesforce’s Agentforce World Tour Sydney on Wednesday to hear about artificial intelligence’s newest era.

Key Takeaways
- 69% of Australian and New Zealand C-suite executives who prioritise AI are focusing on agentic AI over the next 12 months, with 38% already implementing the technology.
- 79% of the 480 executives want to use it to innovate customer experience, give customers a more personalised experience and beat their peers.
- Of leaders who consider AI their top priority, agentic AI was expected to be used for sales forecasting and analytics (58%), 24/7 customer support across channels (68%), platform management (58%), and marketing strategy and campaign development (57%).
- 10,000 people attended Agentforce in Sydney on Wednesday, Salesforce’s annual roadshow to hear about.
What is agentic AI?
Salesforce executive vice president of customer relations management, Bill Patterson, explained to Forbes Australia that while Salesforce has used AI since 2016, it was in the form of a machine-learning system called Einstein, which was ‘making recommendations based on what you knew.’
The last few years of generative AI had been about “predicting based on what you don’t quite know but which is logical and rational,” starting the brief era of co-pilots.
But that era was already finished, Patterson said. “Agents are the new era for the future of enterprise AI computing, because there’s such a shortage of labour in our world, there’s so much shortage of qualified people to actually perform functions.”
He gave the example that while old AI models could answer many basic questions, they couldn’t then sell a product and arrange a delivery driver. And they couldn’t be set up without a knowledge of coding.

A salesperson could set up an agentic AI to scour data looking for leads, then send cold-call emails to potential customers within a certain region. “So what used to take you hours, sometimes days, can now happen in minutes and that’s this amazing moment that we find ourselves in where it’s about the amplification of what every worker can do when they work hand-in-hand with an agent.”
Backstory
With the arrival of large language models in November 2022, Salesforce – the US$292 billion company founded in 1999 – knew it needed to keep pivoting.
“What we were talking about in our upper ranks [at Salesforce] was, ‘Boy if we could simply make the chatbots get smarter, the co-pilots take all of this generative intelligence that we can harness and put into the hands of functions with those same sets of roles and tasks and guardrails, boy we can have a profound impact for our customers,’” recalls Patterson.
“When I talk to CEOs about this, they see this great moment where they’ve been operating with such constraints in the past, this almost becomes a limitless technology.”
Bill Patterson, Salesforce EVP
“When this world of co-pilots came in – we’ve been doing co-pilots for about a year and a half – that was about users starting to ask questions of AI. ‘Can you help me do these tasks?’ We put that in front of customers, and the same questions now started to guide us. ‘Okay, maybe there’s something we could do differently here … So co-pilots gave rise to agents, and agents is now, we think, the future of enterprise AI computing.”
Patterson said 1,000 of Salesforce’s 170,000 corporate customers had taken up Agentforce since it launched last October, while another 4,000 were experimenting with it. It costs $2 per conversation, with a conversation being defined as 24-hours’ worth of interactions.
It was no surprise to Patterson that 69% of c-suite executives were already onto the agentic idea. “CEOs are always trying to think about new ways to become more efficient … Today, most service organisations are overburdened with overwhelming demand from their customers. It’s a sad reality, but most of the time, service teams are answering the same questions over and over again. That monotony leads to a lot of burnout in service teams which leads to a lot of rehiring, retraining and re-fortifying. I think the average rate of exit is like 40% a year. I can now have an agent answer the routine, that means we can use our people for more empathetic services or areas that need more care.

“So when I talk to CEOs about this moment, they see this great moment where they’ve been operating with such constraints in the past, this almost becomes a limitless technology because you can put an agent to do a lot of tasks in the company that you just simply can’t get people to do, and it then amplifies what the humans can do.
“That’s why CEOs know what [AI] agents are, that’s why even before we had the product, they were doing experimentations in this area.”
Look back on the week that was with hand-picked articles from Australia and around the world. Sign up to the Forbes Australia newsletter here or become a member here.