ENG26104. AI Workshop 3.0 - Getting Technical in AI (TCH, FIN)
It is time to move past chatbots and into AI workflows. This advanced, hands-on workshop focuses on designing and executing agent-based AI systems for complex accounting, finance, and advisory use cases. Rather than one-off prompts, participants will learn how to structure AI as digital team members with defined roles, execution logic, and controls.
Attendees will work in small groups on their laptops to design and test AI agents that perform multi-step tasks such as analysis, documentation, review support, and workflow coordination. Exercises will cover defining agent roles using structured instructions, orchestrating agent execution across tasks, and managing context, permissions, and data boundaries. Examples of integrating agents into existing workflows will be demonstrated.
By the end of the workshop, participants will understand how agentic AI differs from traditional generative AI, how to design repeatable AI-driven workflows, and how to apply appropriate control, review, and risk considerations in AI-assisted work.
This is an advanced session intended for attendees already comfortable using generative AI. No programming experience is required, but familiarity with advanced AI concepts is expected.
PRE-REGISTRATION IS REQUIRED. THE FOLLOWING PRE-WORK IS REQUIRED BEFORE ATTENDING THE SESSION: 1) Purchase a license for either ChatGPT, Anthropic Microsoft Copilot or Google Gemini (note: free versions can be used but will have limitations). 2) Bring your laptop to build along with each use case. 3) Download materials that will be provided to registered workshop attendees via separate email. All examples will be done in Chat GPT or Claude.
Learning Objectives:
- Design and configure agent-based AI workflows by defining agent roles, system instructions, execution logic, and context boundaries.
- Orchestrate and evaluate agent execution across workflows, to ensure accuracy, consistency, and alignment.
- Apply governance, control, and risk considerations to agentic AI systems.
