The session’s core message was that European AI scale-ups are winning enterprise customers by solving highly specific operational problems with infrastructure-grade products: ALLSIDES focuses on physically grounded 3D data for physical AI and world models, while Blockbrain focuses on sovereign enterprise AI agents that turn company knowledge into scalable workflows. The available transcript opens with Franz Tschimben presenting ALLSIDES from Brixen, Italy, as a computer vision, machine learning, and computer graphics company building “three d infrastructure” for markets where AI must understand physical objects rather than merely generate plausible images or text.

ALLSIDES’ value proposition is mass production of high-quality 3D data for real-world objects such as tools, toys, household items, fruits, industrial parts, sporting goods, and consumer products. Tschimben emphasized that its scanner can convert an object into a high-quality, physically grounded 3D model in about five minutes, capturing not only visual geometry but also properties such as measured materials, mass, friction coefficients, colliders, and eventually physics. He positioned this as critical for content creation, physical AI, world models, and generative 3D. In product imagery, AI may hallucinate environments or people, but enterprise customers such as Adidas and Nike need the product itself to remain high fidelity and measured. The company said it is already in full rollouts with customers, has around 50 people across Italy and New York, and works with research partners including NVIDIA, the Technical University of Munich, and Heidelberg University.

The ALLSIDES pitch then moved from asset capture into simulation and robot training. Tschimben explained the “real to sim, sim to real” pipeline and the sim-to-real gap: robotics demos often work in simulation but crack in reality. ALLSIDES aims to reduce that gap by producing scarce physical 3D data that cannot simply be scraped from the internet in the way language, image, or video data often can. The company described a 3D data factory in Bressanone scanning sixteen hours a day, six days a week, and announced a San Francisco “real to sim to real factory” with NVIDIA, Industrial Next, and academic partners, using scanners and robots to demonstrate assembly use cases such as server racks and car parts.

Blockbrain followed with Antonius “Tony” Gress positioning the company as a sovereign enterprise AI platform built from Baden-Württemberg for Europe. He argued that Europe’s remaining controllable competitive lever is expert knowledge, especially as Germany faces high energy costs, high labor costs, bureaucracy, taxes, and demographic pressure. Blockbrain’s “brain” turns knowledge hidden in data silos and experts’ heads into institutional memory, connecting to systems such as CRM, SAP, PLM, Google, and Microsoft suites. Its “block” side provides Lego-like building blocks for company GPTs, bots, agents, workflows, always-on agents, and apps. Gress stressed sovereign deployment options, including European servers, European infrastructure, open-source weight models, on-premise and air-gapped deployments, and a close partnership with Schwarz Group’s IT cloud. The provided transcript ends during Blockbrain’s platform architecture discussion and does not include substantive Certivity content.