VietJet Air

Last week I had lunch with a co-worker. Mid-conversation he asked: “So your AI system — could it actually manage a team?” I paused. Because the honest answer is more interesting than yes or no. Over the past few months I’ve built a system that runs cargo analytics for my airline. It ingests reports from various sources and systems, in different formats, detects anomalies, attributes revenue gaps to specific routes and accountable parties, compares against market intelligence, and generates board-ready summaries autonomously. Work that used to take weeks happens in seconds. But here’s what it can’t do: it can’t decide that a partner relationship matters more than the numbers suggest. It can’t read the room when a team member is struggling. It can’t make the call to invest in a route that loses money today but builds strategic position. What it CAN do is make sure I walk into every meeting with complete, accurate, current information — so my human judgment has the best possible foundation. I believe every person in an organisation is a manager of their own work. The title just reflects the scope — from the analyst managing a single report to the CEO managing a portfolio. Peter Drucker wrote that the purpose of information is not knowledge but the ability to take the right action. For sixty years that remained an aspiration. Decision support systems tried and fell short. I think we’re the first generation who can actually deliver on it — not through better dashboards, but through agentic AI that acts on what it sees, not just displays it. The question isn’t whether AI can replace people. It’s whether anyone managing their work — at any level — can afford not to have it.