What DataCrunch does
DataCrunch (now operating as Verda) is a European GPU cloud for AI training and inference. The company positions itself as a vertically integrated, full-stack AI cloud that spans data centers, networking, storage, and a developer-accessible platform, with an emphasis on on-demand provisioning and European data protection and sovereignty. Its platform is described as covering the full AI lifecycle—from rapid prototyping to foundation training and scalable inference—using self-service GPU instances, self-service “Instant Clusters” (including InfiniBand-connected distributed training), serverless containers for inference and batch jobs, plus storage and a container registry. Verda also runs an in-house AI Lab and sponsors/open-support for ML infrastructure projects, which it frames as a way to translate research into platform capabilities. Verda’s customers described on its site include AI startups, enterprises, research institutions, and government/public-sector organizations, and the company highlights enterprise-oriented security/compliance capabilities (for example SOC 2 Type II and additional certifications) and “confidential computing” for privacy-first workloads. Since the DataCrunch rebrand to Verda (executed over 2025), Verda has raised multiple funding rounds, expanded geographically, and reported revenue run-rate milestones in 2026.