Wednesday, November 18, 2026

2:30 PM - 3:30 PM

Bradman Theatre

---

Sponsored by

---

Presentation Outline

To understand the modern battlespace, joint, inter-agency, and multi-national operational systems must fuse an ever-growing set of sources: RF and EW sensor feeds, HUMINT, OSINT, and increasingly space- and cyber-derived intelligence. All this arrives in different formats with different security restrictions. Conventional fusion imposes a heavy data-engineering tax: bespoke pipelines and a pre-built ontology before any operational value is realized, slowing the ability to exploit and make sense of emerging data. I will demonstrate the automation of heterogeneous data ingestion into a multi-model store, enrichment by a governed semantic layer, and orchestration by agentic AI into a confidence-scored common operating picture. Using a representative Indo-Pacific ISR scenario, we show three distinct inputs: a machine-emitted sensor, a human observation report, and an open-source snippet all resolving to a single tracked entity on a map. The fusion is driven by a transparent, source-weighted confidence score providing full provenance to each contributing report. Critically, identity resolution is governed by a curated knowledge model instead of just an AI agent alone. Confidence is computed deterministically from source reliability and independent corroboration (aligned to NATO admiralty grading) keeping a human-auditable evidentiary chain at the centre. Semantic expansion then allows search and query of analyst-relevant concepts whether they are in the source reports or not. Attendees will see how agentic over governed, multi-model data reduces architectural complexity and accelerates onboarding of new source types. This fusion pattern is applicable across ISR, C2, logistics, and multi-domain operations for the ADF and coalition partners.

---

Speaker