ICAEW · 8 MIN
Agents, Context and the Data Foundation - Part 1
Why winning teams build the data foundation before they deploy the agent - and how to start.
Read the article
Founder story
Why I'm the one building this.
I started my career as a chartered accountant. I trained at a firm where month-end could run until 11pm and the partner would still ask why the trial balance didn't tie. I learned, the hard way, what it feels like to be on the receiving end of a number you don't trust - and what it costs to be the person who has to explain it.
The next fifteen years I spent in data and technology consulting - first at the Big Four, then on the client side at HSBC and Barclays, then back in consulting at Accenture. I led data and analytics programmes that ran into eight figures, in environments where Power BI dashboards multiplied faster than anyone could govern them, and where finance teams kept shipping their numbers from spreadsheets - because that was what they could trust.
"The pattern was always the same. The platform shipped. The training happened. The dashboards went live. The business kept running on Excel."
I became a Microsoft MVP in 2020 because I cared about the gap between what the platform could do and what teams actually used it for. I started teaching at conferences - FabCon, SQLBits, Accountex, ICAEW - because the conversations I was having at events were the ones I wished I could have had as a finance leader fifteen years earlier.
I spent two decades on the inside - chartered accountant, product controller, financial modeller, Power BI developer, and now in Avanade's global Centre of Excellence designing how enterprises actually adopt AI. Long enough to watch transformation programmes burn millions and leave the business running on Excel anyway. LDI is the thing I needed back then.
What I am building now is the thing I needed when I was inside: a way for finance and business teams to put Gen AI to work on their actual processes, with the data foundation, the context and the governance that makes it actually work. Self-Service AI is not a product. It is an architecture, a community and a body of practice - and the LDI membership is the way in.