Almost about a year ago I wrote about AI Agent desiderata, essential attributes for autonomous AI systems that can operate responsibly within business contexts. From multimodal understanding and adaptive learning to ethical reasoning and seamless collaboration, these capabilities define the future of agentic AI.

The post explored the potential of systems (agents) that don’t just assist, but act with accountability, resilience, and alignment to human values.

Today, I had the opportunity to present the current state of AI Agents and Agentic AI at the Wellington AI meetup.

You can grab the slides from: AI Agents and Agentic AI – Wellington AI Meetup Presentation.

Why does this matter?
We’re at a turning point where AI is no longer just about automating tasks -it’s about delegating outcomes. Agentic AI systems have the promise on planning, adapting, and acting independently, handling complex connected workflows and making decisions in real time. This shift is shaping how organisations think about transformation, trust, and business value.

Key themes from my talk:

• What makes AI agentic?
  Agentic AI doesn’t just respond—it proactively solves problems, coordinates tools, and adapts to changing conditions, all while maintaining transparency and accountability.
• Business impact:
  We’re moving from “Do what I say” to “Achieve what I need.” Agentic AI enables outcome automation, dynamic adaptation, and scalable intelligence across teams and processes.
• Desiderata for ideal agents:
  Capability, safety, alignment, and reliability are essential. The best agents are autonomous, multimodal, and accountable, with clear guardrails and explainability.
• Real-world examples:
  From customer support and procurement to compliance and onboarding, agentic systems are already reshaping how organisations delegate and operationalise intelligence.
• The design challenge:
  Balancing autonomy, safety, capability, and reliability is key—there’s no universal blueprint, only solutions optimised for specific business contexts.

Personal reflection:
Working with customers in enabling AI as a strategic capability, I’ve seen first-hand how agentic systems can unlock new levels of efficiency and innovation but only when we design for trust, transparency, and alignment with human goals.

If you’re curious about how agentic AI is changing the future of work or want a practical framework for evaluating and deploying these systems, I’d love to have a chat.

AI Agents and Agentic AI – Wellington AI Meetup Presentation.