SirenOpt, a California-based manufacturing technology startup, has raised $25 million in a Series A round led by At One Ventures, with Access Industries, Berkeley Skydeck, Coral Capital, Hitachi Ventures, Visionaries Tomorrow, Voyager Ventures and Wireframe Ventures also participating. The round was announced November 17, 2026, and will fund global sales, applications support and manufacturing expansion for the company's materials-testing platform, according to the company's funding announcement.
Founded in 2022 in San Leandro, California, SirenOpt builds PlasmaSens, a sensing system that combines cold atmospheric plasma — a low-temperature, electrically ionized gas — with physics-informed machine learning to measure multiple physical properties of a material at once, without damaging the part being tested. Each scan captures more than 200,000 data points, which the company's PlasmaSmart software converts into quality readings in place of slower lab-based testing.
Why does measuring materials need plasma and AI?
Traditional quality checks in manufacturing are often destructive: a sample piece gets cut apart, etched or sent to an offline lab, which is slow and tells engineers nothing about every other part on the line. SirenOpt's plasma probe interacts with a material's surface and sub-surface structure without cutting or damaging it, while the physics-informed machine learning model is trained to interpret the resulting signal against known material properties rather than treating it as a black box. CEO Jared O'Leary has said the goal is to put the tool "in every single one of their manufacturing lines, at every single one of their materials manufacturing steps."
What problem is this actually solving for factories?
SirenOpt cites industry estimates that quality-related losses from undetected material defects run 5% to 25% of annual revenue across manufacturing sectors, largely because problems are caught too late in offline testing rather than in real time on the line. The company sells three configurations: PlasmaSens Standalone for off-line analysis, PlasmaSens Motion for piece-by-piece checks, and PlasmaSens Web for continuous in-line monitoring, targeting automotive, aerospace, battery, semiconductor and coating manufacturers across Asia, Europe and North America.
How does this fit into the broader physical-AI and inspection market?
SirenOpt sits at the intersection of two categories that have each separately drawn heavy investment: machine-vision inspection systems, used by companies automating defect detection on assembly lines, and materials-testing instrumentation, traditionally sold by established lab-equipment makers. By combining a physical sensing modality (plasma) with a trained AI model rather than cameras alone, SirenOpt is pitching itself as able to catch defects that vision systems can't see, such as internal material composition issues rather than surface flaws.
The company has not disclosed prior funding rounds, a valuation, or revenue figures, making this Series A the first public marker of its financial scale. The next test will be whether its named manufacturing partners in battery and semiconductor production — sectors under heavy pressure to cut scrap rates — convert pilot deployments into recurring line-wide contracts.
