LLMs in practice: Skills, MCPs, agents and code

Most finance professionals have tried AI to write an email or summarise a report. But using it for day-to-day finance work raises bigger questions: how do you connect it to your numbers, teach it your process, and get results you can trust? In this hands-on workshop, we’ll use Codex desktop to work through a simple finance example together. You’ll learn what agents, MCP connectors and skills actually do, then use them to retrieve practice data, review financial performance and create a presentation that follows your chosen style. We’ll explain each step in plain language, with time to try it yourself. What you’ll learn: How an AI agent works, and how to give it clear instructions How MCP connectors let AI retrieve numbers from a shared data source How to check AI-generated calculations and spot unsupported explanations How to create a reusable skill with your colours, logo and presentation rules How recurring tasks can help repeat a reporting workflow Time to practise with fictional finance data and ask your own questions Before you join Bring your laptop with Codex desktop installed and signed in. We’ll share setup instructions and the practice materials before the session. No coding experience is needed. About the speakerAli is a Finance Content Specialist at Alaan, where he writes about AI and how it's changing finance work. He studied Economics and spent two years in finance before moving into content. He also builds finance tools, writes guides, and runs events for the finance community.