AI gets all the hype – but behind every “smart” feature is a blisteringly fast, brutally efficient platform making it work in production. In this episode of SEEK Bytes, Seamus, Will and Elliott sit down with Ren Shao, a Staff Engineer in SEEK’s Artificial Intelligence Platform Services (AIPS) team, to unpack how SEEK’s AI-driven search really works under the hood – from models and rankings to GeoHash, big-O and CPU caches.

This episode's special guest: Ren Shao (SEEK Staff Engineer)

In this episode, we explore:

• How SEEK’s AI search actually works under the hood – from turning “17-year-old jobs” into normalized queries, to ranking 200K+ jobs in real time, to using historical clicks and impressions to learn what candidates really mean.

• Why this is more “systems engineering” than sci-fi AI – big-O thinking, GeoHash-powered radius search, clever use of CPU caches and HashMaps, and the “latency numbers every programmer should know” that turn textbook algorithms into production-grade platforms.

• How to break into AI platform roles without a PhD – the difference between data engineers, data scientists and “small systems” engineers at AIPS; why curiosity about internals matters more than memorising neural nets; and how open-source code, sites like Codewars and tools like Hugging Face can be your on-ramp.

If you’re in software engineering, data, SRE, platform, architecture or IT leadership and want to see what real-world AI platforms actually look like – not just the buzzwords – this episode is a masterclass in building fast, safe and scalable systems that power millions of job searches every day.

🔔 Follow the SEEK Bytes podcast so you never miss a new episode