Following waves of intelligent automation and Sage Copilot, Sage’s latest focus is on generative AI.

Accounting, payroll and HR vendor Sage has laid out its vision for artificial intelligence at its flagship event, Sage Future, which is being held in Atlanta. Building on nearly a decade of embedded AI automation, Sage led the industry’s adoption of generative AI with its productivity assistant Sage Copilot. It is now positioned to bring the potential of agentic AI and autonomous agents to SMBs in a third wave of innovation.

“What we are calling the three waves of AI over the last several years has become impressive across our portfolio,” said Aaron Harris, Chief Technology Officer at Sage.

Sage’s first wave of task-based AI began its journey by delivering productivity gains with intelligent automation. The early models focused on performing specific tasks reliably, such as anomaly detection and invoice classification. Many customers will have experienced this through features built directly into products like Sage Intacct, where AI improves accuracy and efficiency behind the scenes.

“Sage Copilot was the second stage of this, which makes it easier and more natural to use our software,” Harris said. Through Sage Copilot, customers are using generative AI to discover insights, analyze financial information, and speed up processes like the month-end close. Features like Copilot Close are already delivering measurable time savings and improving confidence in financial data. Sage Copilot is now available to more than 40,000 customers worldwide,

“With the new wave of agentic AI, it can decide on a plan and act autonomously, and will move to where it can act on its own,” Harris added.

Now, Sage is positioned to unlock the potential of the next wave of AI. Unveiling the next phase of its AI strategy, the roadmap shows how Sage is building on its AI heritage, into the generative AI of today and expanding into the emerging space of Agentic AI.

“We are focused on how we do our research, on models that exceed the capabilities of off the shelf models,” Harris noted. “AI is already transforming how small and mid-sized businesses operate, and because automation has been part of our products for nearly a decade, we’re building on a foundation our customers already trust,” he said. “We’re not just delivering generative intelligence, we’re making sure that intelligence scales across workflows, and we’re building toward autonomous capabilities that can support smarter, faster, more confident decision making.”

Harris said that the key in Sage’s model is to choose the right model for the job.

“If what you get off the shelf doesn’t work, we build our own infrastructure,” he stated. “This involves training on accounting with our own large language models.

Harris also indicated that choosing the right model can be difficult.

“No one can agree on what an agent is, so they use specific tools,” he said.

“With the new wave of agentic AI, it can decide on a plan and act autonomously, and will move to where it can act on its own,” Harris said. Sage’s vision for the Agentic AI era for CFOs, is grounded in trust, transparency and customer empowerment .

Agentic AI is all about systems that can coordinate tasks, make reasoned decisions, coordinate tasks, and take actions with minimal direct human interaction. It builds on earlier waves by reducing manual steps and moving toward greater autonomy, while keeping the user in control. This wave will come in two surges. First is Emerging Agentic Applications, Sage’s next generation of AI agents, which will orchestrate complex workflows across systems. In early use cases, customers can ask natural language questions like “What caused the drop in Q2 revenue?” and receive grounded, contextual answers, powered by retrieval-augmented generation (RAG) and multi-agent coordination.

The most advanced wave in Sage’s current strategy focuses on agents with true autonomy: systems that can detect business risks, initiate multi-step responses, and communicate with other agents to resolve issues proactively. These capabilities are already in early-stage testing and are expected to reach customers over the next 12–36 months.

To bring this strategy to life, Sage isn’t just embedding AI into products, it’s transforming how products are built in the first place. By giving developers access to shared frameworks, smart tooling, and built-in guardrails, Sage is accelerating innovation without compromising trust.

Key enablers include:

Sage AI Factory: Secure, self-contained environments that automate AI model training, deployment, and operations for both network-scale and customer-bespoke deployments.  Domain-Specific Models – fine-tuned AI models trained on deep expertise in accounting, payroll, and compliance. Sage Copilot Framework – a development framework that makes it faster and easier to build and deploy Sage Copilot experiences that work seamlessly across products.