The collaboration originated from HydraSpecma’s recognition that AI adoption could not be driven by IT alone. As the Group Strategy Director noted early on: “This isn’t just something IT should take care of.” Instead, the organisation deliberately involved business, HR, and domain experts from the outset — because the value of AI at HydraSpecma lies not in technology, but in how it integrates with the judgment of engineers, specialists, and customer-facing staff.

The organisation faces a structural challenge familiar to many industrial companies: a workforce spanning from production employees who rarely use a PC to digital natives who would pay for AI tools themselves if the company didn’t provide them. With ambitious growth targets and a core identity built on engineering know-how and customer relationships, HydraSpecma’s central question is not whether to adopt AI — but how to do so in a way that strengthens rather than fragments the organisation.

The most distinctive element of HydraSpecma’s approach is their AI Hub — an internal platform on the company intranet where any employee can submit an AI idea, describe the potential time savings, indicate their comfort level with AI, and flag whether they want to be involved in execution. Around 70 ideas have been submitted so far, ranging from meeting note automation to complex engineering workflows.

Behind the hub, a two-dimensional prioritisation matrix evaluates each idea on complexity and business value, routing them into execution tracks. Ideas that are simple and high-value move fast. Ideas that are large and complex get parked until the data foundation is ready. Some ideas submitted as “AI problems” turn out to be basic automation — and get solved that way instead. The process is deliberately unsentimental: if the data isn’t there, the initiative doesn’t proceed in its original form. A dedicated AI Transformation Specialist — an engineer hired specifically because of his passion for AI — owns execution. He builds incrementally, often validating a concept within weeks before deciding whether to invest further, and fills gaps between larger projects with smaller, high-impact tasks like standardising product descriptions across languages.

The most instructive example is Should-Cost, a pricing support tool. The original ambition, feed a 2D drawing into a model and get a price estimate – failed due to insufficient data. Rather than abandoning it, HydraSpecma pivoted: the system now retrieves the 5 most similar past drawings based on ~8–9 parameters. A human expert selects the most applicable and prices accordingly. As one team member noted: “It’s actually an advantage that the expert still has to choose — otherwise we’d just make them lazy.” This is the prediction-judgment split made operational.

A second tool — a BOM and specification checker for wind turbine components — automatically scans bill-of-materials entries against Siemens specifications and flags deviations. It catches errors in a small percentage of cases, but the alternative was shipping non-compliant products to customers. Human review remains the final step. A third initiative in development handles specification update emails, summarising what has actually changed across complex EU standards so engineers don’t have to read every line themselves.

Running in parallel, HydraSpecma has invested in a local AI server to enable experimentation with sensitive documents — contracts, supplier agreements — without routing data through public cloud infrastructure. This reflects a deliberate governance stance: AI should support work, not create new data risks.

HydraSpecma’s model offers a replicable starting point for organisations navigating AI adoption across a mixed-maturity workforce. The AI Hub creates a bottom-up innovation pipeline — surfacing ideas that leadership would never have identified top-down — while the prioritisation matrix ensures resources go where they matter. The dedicated specialist role keeps momentum without requiring every department to develop deep AI expertise.

Crucially, the organisation has chosen to focus AI on freeing up time for human interaction rather than replacing it. As stated explicitly: “We want to preserve the human contact with customers. The goal is to use AI to clear away the time-consuming work behind the scenes — so we have more time to actually listen to what customers need.” This is human-premium thinking applied directly to business strategy. The envisioned outcome is not a more automated HydraSpecma, but a more attentive one — where engineers spend less time on repetitive documentation and more time on the judgment-intensive work that defines the company’s value.