On Wednesday, June 11, we had the pleasure of attending the Refresh 2025 event by Freshworks live.
Hosted in the impressive Emirates Stadium, we were welcomed and immersed in the latest developments, ambitions, and future vision of Freshworks.
One of the most striking innovations in the service management landscape was undoubtedly the evolution of Freddy, the AI companion by Freshworks. This year, Freddy received some major upgrades and was introduced during the event as Agentic AI.
Where Freddy previously acted as an intelligent Q&A system, it is now evolving into an autonomous ecosystem that not only answers questions but also performs tasks independently. Think of handling claims, updating payroll, or booking shipments – all without human intervention.
With these agentic capabilities, Freddy can make decisions, execute actions, analyze root causes, and even suggest next steps. It supports both customers and service agents – not just by answering questions, but by completing actions across multiple applications.
These capabilities are applicable to a wide range of use cases, including order tracking, account changes, bookings, payments, and more.
A key innovation that makes this technology accessible is the introduction of the Freddy AI Agent Studio.
This is a no-code platform that allows AI agents to be easily built, tested, and deployed – without requiring any technical knowledge. Through a visual interface, agents can be taught unique skills, enabling them to autonomously carry out tasks such as checking order status or creating a return request.
Looking ahead: Major evolutions in AI architecture expected
Refresh 2025 made it clear that we’re only at the beginning of a much larger transformation. In the coming months, we expect significant advancements in AI architecture. The focus is shifting from static models to dynamic, context-aware systems that can learn in real time, adapt, and collaborate with other agents within broader digital ecosystems.
We see a clear trend towards:
- Multi-agent systems that communicate and collaborate to autonomously manage complex workflows.
- Contextual AI, where agents don’t just respond to input, but anticipate needs based on historical data, behavioral analysis, and external signals.
- Federated and edge AI architectures, enabling faster processing, improved privacy, and scalability across platforms and devices.
- AI governance and ethics by design, with transparency, accountability, and responsible AI use embedded in the architecture.
These developments won’t just change how businesses adopt AI – they promise to accelerate a shift that will position AI as a strategic core component of modern organizations.
