OpenAI's EMEA FDE head Eoin Boylan: rare fine-tuning, Codex-based builds, and two-engineer pods doing what once took six

Towards AI's weekly newsletter (TAI #228, Oct 30) includes a conversation with Eoin Boylan, OpenAI's Head of Forward Deployed Engineering for EMEA. He says his FDE team concentrates on model intelligence gaps and missing product capabilities. OpenAI's Deployment Company handles rolling out existing capabilities at scale, and partners bring industry relationships. Some of his FDEs work inside research to close capability gaps found at customers.

Every engagement starts with discovery that tests whether a use case will pay off and can be justified to the customer's board. At one semiconductor customer the team listed more than 50 use cases and ranked them by complexity and value, starting with quick wins to win over the hardware experts. Boylan says the team now rarely fine-tunes for individual tasks, because base-model improvements keep overtaking that work. Most solutions are built on the Codex harness with plugins and skills. He also describes building custom features, only for the product to ship the same thing a month later.

For FDEs, the practical signals are smaller delivery teams and new kinds of work. In one example, scope that used to need six engineers and a technical delivery lead is increasingly handled by two engineers and a lead. Engineers define domain boundaries, interfaces and tests, then let agents implement within them. Legacy modernization, such as moving banking systems from COBOL to Java, is a large part of the work. One example is a Spanish bank, where the team is cutting the time analysts spend on compliance requirements for commercial credit memos.

Towards AI, which runs its own FDE consultancy in pods of two AI engineers/FDEs and one deployment strategist, says it is hiring FDEs, deployment strategists and a platform and security lead in London. The newsletter also notes that OpenAI is hiring FDEs in London.