Producing Capability: A Leader's Guide to Rapid, Safe Prototyping with Harshil

For more details visit the Meetup page

The call for a more productive, capable public service is clear. But how do we move from "AI concept" to "practical capability” without the high-risk, multi-year project cycle? This session is a leader's guide to doing it. This is a real conversation between Cheng Yu (Head of R&D, Trellis Data) and Harshil Siyani, an AI developer in the Innovation Department at News Corp Australia and a key member of the Canberra innovation community. His day-to-day work is applied AI and rapid prototyping, taking unknown or ambiguous ideas and converting them into working capability quickly. He has previously presented at ACS and Google Developer Groups (Melbourne and Canberra), and has been featured in national print coverage for his analysis on bias in AI. Harshil will talk through examples like structured data problems where AI-to-SQL wasn’t enough, and over-engineered prototypes that didn’t need to be complex. He’ll show how rapid prototyping reveals the cleanest, lowest-friction path so policy outcomes can be tested and validated faster and safer, before money gets burnt. This is the new way to build: faster, safer, and genuinely innovative. You'll learn how to test, validate, and de-risk your AI initiatives before you commit. This briefing is for the leaders, policymakers, and builders who want to see behind-the-scene process, and learn how to control it.