Most people will now have had some exposure to Large Language Model tools such as ChatGPT and will have been impressed by their ability to write essays or generate computer code. However, fewer people will be aware of the next phase of AI, which will turn AI chat assistants into more fully fledged AI colleagues. These emerging set of tools come under the broad term of “AI agentsâ€, which are AI-based systems which have capabilities beyond a simple chat interface and are able to interact with external tools, access data, and most importantly have the capability to plan and perform arbitrarily complex tasks. In this talk we will explore the emerging field of AI agents and discuss our experiences using such frameworks to build end-to-end actuarial modelling solutions and demo some POCs we have been working on for the audience. We will also cover topics related to building such complex AI systems including: prompt-engineering fine-tuning retrieval augmented generation function calling specific AI agent frameworks and how they will allow AI to participate in all the different stages of the actuarial modelling process from coding and testing to documentation and analysis.