GramEye lands ¥580M NEDO grant to take AI bacteria diagnostics to the US
What's the deal? Japan's GramEye has been re-selected for the PCA phase of the New Energy and Industrial Technology Development Organization's (NEDODealroom has a profile for this one. Try Dealroom →) deep-tech startup support programme, securing a grant of roughly ¥580 million. The Osaka-based company uses AI and robotics to modernise Gram staining, a common test for identifying bacteria and guiding antibiotic choice.
By the numbers: The grant covers eligible costs of about ¥870 million and lifts GramEye's total funding to ¥2.25 billion, including ¥1.23 billion in grants. As a grant, it sits in the 37th percentile by size — modest against the wider funding pool, but a meaningful vote of confidence from a national research agency.
What's the endgame? GramEye wants to become what it calls the "flagship company" for solving antimicrobial resistance (AMR), which it frames as a public-health challenge on par with climate change. The company will use the funds to develop a next-generation machine for the US mainstream market, pursue FDA clearance, and build an AI platform for microbial identification.
The product: GramEye launched its Gram-stain automation system, Mycrium, in January 2025. The tool automates a staining process long considered difficult to mechanise and uses AI to support microscopy, aiming to deliver faster, more consistent results and cut the reliance on technicians' subjective judgment.
Why now? Gram staining is often skipped at nights and weekends even in large hospitals because it demands time and specialist expertise. GramEye argues automation can standardise the test and promote appropriate antibiotic use — pressing as drug-resistant bacteria spread. The company cites a widely referenced 2014 review projecting that, unchecked, AMR could cause 10 million deaths a year by 2050, exceeding cancer.
What could go wrong? The plan hinges on entering the US market, where FDA clearance is a hard gate for medical devices. Chief executive officer Yu Hiraoka calls the AMR problem "an urgent challenge that cannot wait" (translated from Japanese), but regulatory approval and building a US commercial footprint remain unproven for the company.
The signal: NEDO's PCA phase is designed to back deep-tech firms moving from prototype to market. Re-selecting GramEye for later-stage commercialisation support signals Japan's interest in exporting home-grown medical AI — and in positioning AMR diagnostics as a global growth market.
Image credit: GramEye
