In this talk, I’ll share some of my failures as an engineer navigating the complexities of identity standards—highlighting the mistakes, challenges, and lessons learned along the way. Writing code to present standards information differently has been a game-changer in how I absorb and apply knowledge, and I’ll explain why this approach works for me.

We’ll explore the journey I took by going back to basics—starting with data. I’ll explain why large language models [LLMs] aren’t yet up to the task of learning identity standards. I’ll also introduce the IANA registries as the unsung hero that’s essential for organizing the vast array of identity-related data, and how tools built to leverage this data have transformed my understanding.

Finally, I’ll discuss the difficulty of capturing and translating intent and spirit in specifications, and why this remains one of the hardest challenges in implementing identity standards accurately. By the end, you'll have a better grasp of the real-world struggles behind identity standards, and how you can tackle these challenges by thinking differently and embracing new tools and approaches.