Quantum Funds Support Startup Quminex in Developing AI System

On October 14, the startup Quminex from Sherbrooke, Quebec, closed a pre-seed round of 1.5 million Canadian dollars (~$1.06 million). The deal was led by the Quantacet fund, joined by Swiss QAI Ventures and business angels. The funds will be directed towards the Quminex Discovery Platform – a service for selecting drilling locations. The company has already completed its first commercial projects: identifying targets on plots for Tech-X Resources from British Columbia and an unnamed mid-sized mining company. Contract amounts are not disclosed. Quminex will also use the platform for itself – to purchase exploration licenses in Canada.

According to QAI Ventures CEO Alexandra Beckstein, at the early stage, geologists still rely on expert judgment: data is scarce and fragmented. Financing is declining – according to S&P Global, in 2025, budgets for finding new sites decreased by 8% to $2.57 billion, which is 21% of global spending on subsurface exploration in the non-ferrous metals segment, a record low share.

The platform turns this work into a reproducible computational process: it accepts geochemical, geophysical, and spatial data from the client, cleans it, standardizes it, constructs features, and selects the most informative ones. Machine learning algorithms are trained on them. The result is an explainable mineral system model that shows which factors influenced the forecast. The client receives a list of targets. The company does not disclose accuracy metrics.

Quminex CEO and co-founder Mark Fassbender studies rocks hosting sulfide deposits. In April 2025, the startup described the technology as a combination of AI and quantum computing, but in the new release, the second component is not directly mentioned, and on LinkedIn, the company presented it as HPC. The team includes two specialists in the quantum field, one of whom is studying how quantum algorithms can improve predictive models, although they have not yet shown practical advantages over classical computers in any applied area.