Founders: Ashin Alex, Sam Kurian and Jimin Seo

Insurers, miners, and farmers are sitting on mountains of map-based data made up of satellite imagery, drone footage, and radar data. But almost none of it is in a format that AI can use. The Auckland spatial-AI lab Hyades reckons it can fix that, and has raised a NZ$1.1 million (A$910k) pre-seed round.

The round was led by Icehouse Ventures, with backing from K1W1 and angel investors Tony Falkenstein and Tim Brown. Hyades also secured an NZ$400,000 government "New to R&D" grant.

The platform, called HyperPrism, is still in early alpha and aims to fuse fragmented, messy spatial data into a single, AI-ready format for industries such as insurance, agriculture, mining, and climate science. The flagship use case is risk modelling. Today, an insurer wanting to gauge future flood exposure must do the slow grind of manually collating imagery, flood records and other data. Hyades automates it, letting them quickly build models showing which properties face the greatest risk of flooding, cyclones, or wildfires.

Its other product, BlazerOps, allows users to build partial machine learning pipelines and optimised architectures for individual companies.

Founded by University of Auckland grads Ashin Alex, Sam Kurian, and Jimin Seo, Hyades will spend the cash on enterprise co-design partners, AI engineering hires, and scaling toward a wider release.