Ten years next month — October 2015 — since the academic-side conversation that produced the 2016 February intern hire that ultimately produced the EmilyAI work. The intern was, at the time, recommended by an academic contact with a research interest adjacent to where the company's SOC work was already heading. The first conversation was a brief one in a small London cafe near the offices of the time. The decision to make the hire was not, at the moment, framed as a strategic-product-development decision; it was framed as a useful research-experiment hire with a small initial budget and a six-month working-window.
The decade since has been the principal product-development arc of the company. The shadow-mode initial deployment in 2016. The June 2016 first writing about the work in this blog. The internal-deployment maturity through 2017. The customer-pilot in April 2018. The ongoing customer-acquisition trajectory through 2019-2025. The successive product-capability additions — threat-intelligence-integration, multi-SIEM support, language-model-integration, federated-learning. The team-growth around the work from one intern to twelve full-time engineers and customer-success staff. The book-and-conference-contribution arc — USENIX Security 2019 paper, Black Hat USA 2021 paper, USENIX Security 2024 paper.
The ten-year-arc would not have been predictable from the cafe-conversation in October 2015. The hires-that-shaped-the-company are not consistently predictable in advance; the operational-judgment-call to make the hire and the subsequent operational-judgment to support the work as it developed are the substantive contribution that the company-leadership-position produces. I am, on the personal-reflection measure, content with the choices made along the arc.
The October 2025 ten-year-of-the-arc marker will be unmarked. The lead engineer's continuing work continues. The arc continues.
The blog continues.