How Stockland Is Becoming an Agentic Enterprise: Lessons From Its Chief Customer and Digital Officer

Stockland is using AI agents to give its people time back, starting with sales, service and complaints, and redesigning work around outcomes rather than existing processes. Chief Customer and Digital Officer Sharmila Tsourdalakis shares what Australian leaders can learn.
Stockland is one of Australia’s largest diversified property groups, spanning retail, residential, logistics and investment management. It’s also a Salesforce customer that is now putting AI agents to work across the business.
Sharmila Tsourdalakis, Stockland’s Chief Customer and Digital Officer, has spent more than 20 years leading technology, innovation and digital transformation, and she is now driving Stockland’s move into agentic AI. In a recent conversation with Leandro Perez, she explained where Stockland started, what it has learned along the way, and where it’s heading next.
Key Takeaways
What does “AI-native” mean for a diversified property group?
For Stockland, the starting point wasn’t the technology; it was the business strategy.
“We have a diversified portfolio strategy. We’re known as both a developer and an investment manager, and we have a very clear pathway for growth across many asset classes,” Sharmila says. AI is changing the shape of real estate, so Stockland’s focus is on what it means for people, communities and the growth of Australia.
The second question is how to scale. “What does that mean for how we can scale our business to deliver this next phase of growth in ways that are different to how we’ve imagined, that can mean something beneficial for our people and for communities?”
Sharmila doesn’t see a fixed gap between where Stockland is and where it wants to be. “While we are in this conversation, something is changing on AI.” The mission, “for people, for communities, for something better”, gives everyone a reason to keep reimagining. Her CEO calls it a beginner’s mindset.
Why did Stockland start with internal agents?
Stockland has been a Salesforce customer for about 12 years and now runs multiple agents powered by Agentforce that support its internal teams.
The reasoning was straightforward. Last financial year, Stockland doubled its volume of residential settlements compared with the year before. “If we’ve got a strategy around growth, how can AI allow us to scale in a different way that allows our people to do more higher-value work?” Sharmila says. “We don’t want our people needing to do more of the same things as we grow.”
So Stockland deliberately started internally, where it could give capacity back “in ways that are trusted”, learn quickly and bring its people along. It looked at enquiry volumes, service cases and areas where customers might be at risk.
Sponsorship is shared. The work is led by the whole senior leadership team with strong backing from the CEO, and Sharmila runs it “hand in glove” with the Chief People Officer.
What do Stockland’s AI agents actually do?
Three early use cases show how agents sit inside the flow of work in Salesforce:
- Settlement risk. Sales teams work in Salesforce with an agent that flags risk points for customers who are heading towards settlement, so the team can support them proactively.
- Knowledge at their fingertips. Stockland receives thousands of enquiries. Agents surface the right knowledge so people don’t have to jump from screen to screen.
- Faster, more authentic complaint handling. Agents use data already in Salesforce to give automated insights into what needs action, “before we’re waiting for that phone call.”
The aim is to free people for what Stockland calls its customer strategy’s “moments that matter”. Those are the points in a customer’s journey where a human connection counts most.
Lesson 1: Let your people shape the agenda
Stockland wanted to “show, as opposed to tell” how AI creates value. The team started with the biggest pain points in the organisation, using employee survey feedback and conversations with frontline teams who know where the friction is.
That led to three pillars:
- Employee experience, including a procurement process that benefited from unlocking data.
- Customer and sales, developed in close partnership with the sales team, who wanted help to close more deals and accelerate their work.
- A vision for the system of agency and engagement, meaning how agents and people will work together.
“It was a mix of both,” Sharmila says: frontline engagement and executive focus on the value that could be created.
Lesson 2: Unlocking knowledge is the real work
Many organisations expect an agentic journey to be mostly about technology. Sharmila says Stockland expected the same. “The tech leaders tend to do that.”
What took the most time was something else. “It’s unlocking knowledge, all of those things that are in the minds of our great people, and then being able to redesign how we work, learn from it and experiment.”
That includes how procedures are documented. Stockland, like most organisations, has lots of documented procedures, but capturing knowledge for an agent is different from capturing it for a person. Stockland keeps a human in the loop and adapts how its knowledge is recorded so it suits agent engagement.
Lesson 3: Redesign work from the outcome back
Stockland’s early agents began with use cases. Now it’s shifting to redesigning work end to end, focused on the outcome rather than department to department.
“If this is the outcome I want, be it greater stickiness in the customer experience or further uplift in net promoter scores, and you redesign from that back, you may come up with a very different answer,” Sharmila says. She describes it as moving “left to right as opposed to right to left”.
Her first phase was an introduction to AI, learning what models can do. The second is about how to scale and industrialise. And if you lean into it with a people and outcome focus, “it’s actually not a refinement of process. You can do something foundationally that is going to revolutionise what you’re doing.”
Because the pace of change is so fast, Stockland regularly stops and asks: what is the outcome, and what are the few levers that matter? It’s also comfortable changing course. “Sometimes you’ll have an agent, and three months later, things have shifted so significantly that it’s OK to change.”
Lesson 4: Bring people along with a steady cadence
“Our people truly are at the heart of this,” Sharmila says. Leaders set the direction and make sure it’s grounded in strategy. Then they free people to reimagine without being limited by past constraints.
Over the last nine months, Stockland has run a strong cadence of engagement:
- An introduction to unlocking value through AI.
- A program for every leader on the art of the possible and the questions to ask.
- A session a few months ago with more than 250 people working together on what she calls a “vibe coding session”. The aim was to imagine new solutions from the outcome you want, rather than from requirements and documentation.
Leaders also play a role by showing how they use AI. “I let my team know, hey, that’s something you would have had to do that I’m now doing for you, just letting you know,” she says. When people see leaders being curious, “it creates a ripple that flows through the organisation.”
So why does an Agentic Enterprise need Salesforce data?
Sharmila is candid that nobody knows exactly where AI is heading. But she’s clear about what she needs today. “You need to be able to record information. You’ve got to ensure that it’s captured, and therefore you do need that combination of probabilistic and deterministic.”
In practice, that means pairing AI models with reliable, governed records like those in Salesforce. “The real power is in how this interacts seamlessly within the wider ecosystem. If you can unlock the data, unlock context and knowledge, and tap into the choice of models that are there, it’s going to create new value pools that haven’t existed before.”
What’s next for Stockland?
Next year Stockland celebrates 75 years of heritage. Its mission, for people, for communities, for something better, shapes where it’s heading.
One big opportunity is the residential home journey. “We know that means a lot to Australians,” Sharmila says. Stockland wants to take what it has learned internally and see where it can create value for customers: easing the burden and simplifying a journey that involves many parties, particularly for first-home buyers doing it all for the first time.
“We’re looking deeply at what we can do next to make the Australian dream easier to achieve, in ways that agentic capability can really support our customers.”
What Australian leaders can take from Stockland
- Anchor to your growth strategy. Ask how AI can help you scale, not just where it can be applied.
- Start internally. Build trust and learn with agents that support your own people first.
- Listen to the frontline. Use survey feedback and conversations to find the biggest pain points.
- Invest in knowledge. Capture expertise and structure it so agents can use it, with humans in the loop.
- Work back from the outcome. Redesign end to end instead of automating the current process.
- Keep engaging your people. Train, show leadership curiosity and give people room to experiment.
Frequently asked questions
An agentic enterprise is an organisation where AI agents work alongside employees to handle tasks and surface insights, supported by connected data and redesigned ways of working. Stockland is an example, with multiple Agentforce agents supporting its sales, service and complaints teams.
Start where your strategy and your people’s pain points meet. Stockland began with internal use cases in sales, service enquiries and complaints, tied to its growth plan, so that agents could give capacity back to its people in a trusted way.
Start from the outcome you want, such as higher customer satisfaction or net promoter score, and work back, rather than adding an agent to the current process. Stockland also involves teams end to end instead of department by department.
Combine regular engagement, training, leadership role-modelling and hands-on experimentation. Stockland ran an introduction to AI, a program for every leader and a session with more than 250 people imagining new solutions from the outcome they wanted.
No. Organisations need to record and capture information reliably, which means combining probabilistic AI with deterministic data such as the records in Salesforce.
Ready to build your own agentic enterprise?
Explore how Agentforce and Data 360 help Australian and New Zealand organisations put AI agents to work in the flow of work. Read more customer stories from across ANZ.
Watch more Executive Conversations







