OpenAI and AI Singapore run their first FDE bootcamp, and it exposes a talent gap

AI Singapore and OpenAI have finished the first OpenAI for Singapore FDE Bootcamp, a three-day, hands-on program co-organised by the two. It is part of OpenAI for Singapore, the multi-year commitment of more than S$300 million announced in May, which includes over 200 technical roles, OpenAI's first Applied AI Lab outside the US, and a forward deployed engineer training programme for local talent.

In an essay published on 10 October, Laurence Liew, who runs AI Singapore's AI Apprenticeship Programme (AIAP), says his engineers met the standard expected of them, but names three gaps between his graduates and what OpenAI's Singapore FDE listing asks for: years of production software engineering (the listing wants at least five), hands-on work with the frontier stack such as evals, agents and retrieval, and the habit of owning a customer outcome alone rather than inside a mentored team.

He plans to add more frontier-stack engineering to the program and give apprentices solo customer ownership earlier, while being upfront with hiring managers that a nine-month program can't replace five years of production experience.

Why it matters for FDEs: frontier labs are now hiring FDEs outside the US and building local training pipelines, and this is a candid account of what they actually screen for. If you're aiming for an FDE role, production engineering depth and real customer ownership count for more than AI coursework alone.