More than 2,500 people are expected at RAI Amsterdam when HumanX Amsterdam opens on September 22, and The San Francisco Tribune’s preview of the companies worth watching has already narrowed the field to nine names drawing early attention. The event is part of the broader HumanX Europe circuit, and organizers say 60 percent of attendees will hold titles at the VP level or above, a signal that the conversations happening on the show floor will lean heavily toward deployment rather than demonstration.
Data Access and Governance
Two companies at the event are tackling what happens before AI can even get to work: getting it access to the right systems, and keeping that access accountable. Unfold, still operating in stealth, is building a way to read closed, undocumented, or vendor-locked enterprise systems without requiring an API, documentation, or a vendor’s cooperation. The company’s approach runs inside a customer’s own environment, with read-only access by default and full auditing built in. Its position is straightforward: if the business logic inside a system stays locked away, modernization can’t replace it and AI can’t reason over it.
Trendium approaches the same territory from the governance side. Its ContextGuard platform gives enterprises visibility and control over their AI infrastructure, and its AI Enablement Program is built to help engineering teams use AI coding agents responsibly. The company organizes its work around four pillars: visibility, control, enablement, and compliance, a framework that speaks to a shift in enterprise priorities. Speed alone is no longer the goal; understanding what an AI system is touching and doing has become just as important.
Making AI Reliable at Scale
causaLens is working on the gap between a promising agentic demo and a system that can actually be trusted in production. Its Digital Knowledge Workers use multi-agent architectures to automate repetitive knowledge work, built on a platform that combines pre-built Blueprints, a Factory for assembling customized workers, and a governance layer called the System of Work. Causal reasoning, human-in-the-loop controls, automated evaluation, and self-healing are all part of the pitch, aimed at producing outcomes that hold up outside a controlled setting.
Kensho Technologies operates as S&P Global’s innovation engine, pairing AI and engineering talent with the company’s financial and business data. Its technology connects large language models, agents, and AI applications to trusted S&P Global information, while also structuring proprietary data for machine learning and generative AI use. For finance and other data-heavy industries, that combination of domain expertise and infrastructure represents a meaningful piece of the enterprise AI stack taking shape right now.
Voice Becomes Infrastructure
Three companies at HumanX Amsterdam are building the layer beneath voice AI. AssemblyAI provides speech-to-text and voice infrastructure through APIs that support transcription, contextual understanding, and real-time agentic workflows, letting developers add voice capability without building it from scratch. Speechmatics focuses on low-latency recognition for multilingual, multi-speaker settings, supporting more than 55 languages and deployable in the cloud, on-premises, or directly on devices, serving use cases from healthcare and legal transcription to contact centers and live captioning.
Otter AI is pushing past transcription entirely, describing its direction as a conversational knowledge engine. The platform records and transcribes meetings, flags decisions and action items, and makes past conversations searchable, with users able to query accumulated meeting knowledge, generate follow-ups, and connect information to workflows like CRM systems.
Discovery and Commerce
Nuritas applies AI to peptide discovery through its proprietary Nuritas Magnifier platform, which has already identified more than 8 million peptides and built a substantial library of known functionalities. The company moves discoveries from computational prediction through clinical validation, shortening a process for health and nutrition ingredients that traditionally takes years. PhotoRoom, meanwhile, is building AI visual infrastructure for e-commerce, offering batch editing, automated quality assurance, and brand controls for sellers ranging from individuals to large enterprises managing millions of product images, with a particular emphasis on keeping generated images faithful to the actual products being sold.
The Bigger Picture
Taken together, these nine companies point to where enterprise AI is actually headed: not toward flashier models, but toward the infrastructure that makes existing ones usable at scale. HumanX’s own programming reflects that shift, with dedicated tracks like VentureConnect and SolutionBridge built specifically to connect startups with enterprise buyers and investors. For anyone attending, the companies worth a longer conversation may be the ones solving problems that never show up in a product demo.
