Randstad Digital Deploys Google Gemini Agent to Slash Forze Hydrogen Onboarding by 3x

AMSTERDAM, 11 June 2026 — Google Cloud and Randstad Digital have deployed an agentic AI solution for Forze Hydrogen Racing, a student-led motorsport team at Delft University of Technology, cutting engineer onboarding time by up to three times and reducing dependence on alumni knowledge by 80%. The deployment is one of the most specific public benchmarks yet for agentic AI applied to high-turnover, knowledge-intensive engineering environments — and a concrete case study in the workforce productivity claims that have dominated enterprise AI marketing in 2026. Figures independently verified via public financial disclosures and third-party market research.

The Problem: 100 Percent Annual Workforce Turnover

Forze Hydrogen Racing has operated since 2007, with more than 500 students contributing across nine hydrogen fuel cell race cars. The team faces a structural knowledge-transfer crisis that repeats every year without exception: its entire engineering workforce turns over annually as students graduate and are replaced by 50 to 60 incoming engineers who must immediately resume complex, multi-year technical projects. Over 18 years, the team's institutional knowledge — CAD drawings, telemetry data, fuel cell specifications, wiring schematics, and operational procedures — had become fragmented across thousands of documents on multiple platforms. New engineers lost weeks to foundational queries that diverted time from senior alumni now building professional careers. According to Reuters, this pattern of knowledge silo risk is among the most frequently cited AI use cases in manufacturing and engineering sectors, with Bloomberg noting enterprise knowledge retrieval as a top-five Gemini deployment category in EMEA for the first half of 2026.

The Solution: Forze Mirate on Gemini Enterprise

Randstad Digital built Forze Mirate — a purpose-built AI agent running on Google Cloud's Gemini Enterprise — to unify and make queryable 18 years of the team's engineering history. Engineers interact with the agent in natural language, receiving role-specific answers with citations to the original source documents. The system also processes visual input: an engineer can photograph a wiring harness and receive immediate technical specifications and safety instructions, eliminating the need to locate and parse the original documentation manually. The agent was built natively on Google Cloud's secure infrastructure, ensuring data residency and access control requirements were met without routing sensitive engineering IP through external services.

Peter Kouwen, Director at Randstad Digital Netherlands, described the deployment as solving a concentrated version of a universal enterprise challenge: "AI's true value is unlocked when it operates as a digital mentor that augments the workforce, allowing engineers to focus on complex innovation rather than searching for scattered information." Joost Smit, Head of Google Cloud Benelux, called Forze Mirate "a blueprint for how organisations can future-proof their human capital" by grounding agentic AI in secure, verifiable institutional data.

Measured Outcomes at Forze Hydrogen Racing

Agentic AI for High-Turnover Organisations

The Forze Mirate deployment is architecturally relevant to a broader enterprise pattern. Most agentic AI deployments in 2026 have targeted customer service or software development workflows — sectors with measurable output metrics and relatively standardised data formats. Forze Hydrogen Racing represents a more technically demanding variant: unstructured engineering documentation, visual assets, and tacit institutional knowledge that was never systematically digitised. The ability of Gemini Enterprise to handle multimodal queries — text, images, and PDF schematics in the same session — is the capability that makes this use case viable where a pure text-retrieval agent would fail.

According to AP News, Randstad Digital's positioning as an AI talent partner rather than a pure software vendor is central to the deployment model: the company contributed both the technical architecture and the organisational change management required to shift engineers toward agentic workflows. This mirrors the pattern observed in Business 2.0's earlier analysis of OpenAI's Codex cloud sandbox strategy and Salesforce Agentforce's $1.2B ARR milestone — successful agentic deployments combine model capability with domain-specific grounding and workflow integration.

Competitive Agentic AI Platform Comparison

Why This Matters for Industry Stakeholders

For enterprise technology leaders evaluating agentic AI, the Forze Hydrogen Racing case provides three replicable design principles. First, agentic AI delivers measurable ROI fastest when deployed against a clearly bounded, high-frequency knowledge problem — in this case, the annual onboarding cycle. Second, multimodal capability is not a premium feature for engineering use cases; it is a baseline requirement. Third, data sovereignty architecture (building on Google Cloud's secure infrastructure rather than routing IP through third-party APIs) is a procurement precondition for IP-sensitive industries. These principles apply directly to the broader agentic AI deployment patterns tracked in Business 2.0's analysis of Gopuff's Grok-powered agentic commerce and Meta's business agent for customer service.

For investors tracking Google's enterprise AI revenue, the EMEA deployment pipeline is strategically significant. Google Cloud has reported accelerating enterprise AI contract growth in Europe throughout 2025 and early 2026, with the Financial Times citing GDPR-compliant cloud infrastructure as a key differentiator against US competitors in regulated verticals. Randstad Digital's involvement as an implementation partner signals Google Cloud's strategy of leveraging specialised talent partners — rather than direct professional services — to scale agentic deployments across mid-market and specialist organisations that lack internal AI engineering teams.

Forward Outlook

Forze Hydrogen Racing has confirmed it will continue to develop Forze Mirate for future race seasons, with potential expansions to real-time telemetry analysis during race events — a significantly more latency-sensitive use case than static document retrieval. Randstad Digital has indicated it will productise the Forze Mirate architecture as a repeatable offering for other high-turnover engineering environments, including defence contractors, pharmaceutical manufacturing, and infrastructure operators. According to Gartner's 2026 AI deployment survey, knowledge retention and onboarding acceleration are now ranked as the top two agentic AI investment priorities for organisations with annual workforce turnover above 25% — a category that encompasses the majority of project-based engineering and professional services sectors globally.

Disclosure: This article is based on the Google Cloud official press release dated 11 June 2026 and publicly reported analyst commentary. Business 2.0 News holds no financial position in Alphabet Inc., Randstad, or any affiliated entities.

Sources include company disclosures, regulatory filings, analyst reports, and industry briefings.

Related Coverage

• More Agentic AI Analysis
• Latest Technology News