Government holds some of the most valuable data in the country, and AI is raising the stakes on how it's managed, understood and trusted. Hear from government and industry leaders on how agencies are making sense of complex data and what it takes to adopt AI responsibly and on their own terms.
With breakfast and brunch, come join us for a morning of learning, connection, and discussion around graphs and AI in government.
Why attend:
Customer story: NFSA. See how the National Film and Sound Archive of Australia are using emerging technologies, including AI and graph, to proactively preserve and shape knowledge in a rapidly changing digital world.
Sovereign AI in government. Hear from Mistral AI on what sovereign AI means for the public sector and how agencies can adopt it responsibly.
Live demo: See how an AI agent, grounded in a Neo4j knowledge graph, can assess regulatory amendments in minutes—rapidly connecting each change to the obligations, controls, processes, owners and projects it affects
Latest from Neo4j: Get the latest updates and announcements from Neo4j.
Dr Keir Winesmith, Chief Digital Officer, National Film and Sound Archive of Australia
Today there are a range of disruptive forces impacting cultural and memory institutions. First among those is the rise and prevalence of digital technologies and their use in the creation and consumption of culture. Keir Winesmith has been exploring new technologies, including machine learning and graph, for over two decades.
Through a set of lived examples and speculative provocations, this talk will sketch how we might respond proactively to this moment
From Connected Knowledge to Sovereign AI - Building government capability we can inspect, test and control
Andrew Goodall, Australian Government Partnerships, Mistral
Government does not need another AI demonstration that produces a convincing answer. It needs capabilities that work with institutional knowledge, preserve evidence and uncertainty, and remain under accountable control.
Drawing on international experience, this talk explores how connected knowledge, domain-adapted models and mission-specific evaluations can work together in high-consequence environments. We will examine where retrieval helps, where model customisation can add value, and why neither a powerful model nor a knowledge graph provides assurance on its own.
The practical starting point is not a perfect enterprise platform. It is a strategically important, bounded use case—with executive sponsorship, clear operating boundaries and evidence of fitness for purpose. For defence, intelligence, law enforcement and federal government teams, the objective is capability they can inspect, govern, improve and sustain.
