From Microscope to Machine Learning: How Techion & HSO Turn Parasite Testing into Real-Time Insight

Techion Group is a New Zealand ag-tech innovator best known for FECPAK, a portable digital-microscopy platform that lets farmers capture faecal egg-count (FEC) images on-farm and send them to the cloud for analysis and long-term tracking. Since 2011 the company has pioneered “microscope-in-the-cloud” services, building the world’s only global parasite-disease database and processing tens of thousands of samples every year.

Rapid success created a new constraint: every extra country, species and customer required more technicians to label images around the clock. Microsoft analysts estimated that undetected drench resistance already costs NZ farmers NZ$98 million a year, yet Techion’s turnaround times could still stretch to days, pushing producers to treat animals pre-emptively.

Internally the firm was sitting on more than 120 000 expertly annotated images, but off-the-shelf computer-vision models topped out at only 47 percent accuracy on the tiny, crowded egg images — far below diagnostic standards.

Techion evaluated several AI vendors but chose HSO for its deep Microsoft expertise, proven computer-vision track record and collaborative approach. HSO’s architects helped balance global model coverage with local sample bias and set up one-click retraining pipelines so Techion’s scientists can fine-tune models as biological conditions evolve.

Building on the Microvision core, Techion is partnering with Awanui Labs to adapt the technology for human diagnostics, water quality monitoring and broader environmental testing, aiming to bring point-of-care microscopy to underserved regions. Parallel advances in low-cost digital microscopes and AI-assisted parasite research underscore the vast potential to automate microscopic workflows well beyond agriculture.