How DHL Built an AI Skills Marketplace for 570,000 Employees — Including the Ones Who Never Sit at a Desk
How DHL Built an AI Skills Marketplace for 570,000 Employees — Including the Ones Who Never Sit at a Desk
Most enterprise AI-in-HR stories follow the same script: a global company deploys a skills platform for its knowledge workers, publishes impressive adoption numbers, and calls it a transformation. Deutsche Post DHL Group did something harder. It built an AI-powered skills marketplace that works for parcel handlers and delivery drivers — not just the people with laptops.
That distinction matters. DHL employs roughly 570,000 people across more than 220 countries, making it the world's largest logistics company by headcount. About 400,000 of those employees are field workers — drivers, warehouse operatives, sorters — whose roles look nothing like the office-based positions that most talent platforms are designed around. Any skills taxonomy that ignored them would cover barely a quarter of the workforce.
The Problem: Two Workforces, Zero Unified Skills Language
Before AI entered the picture, DHL's learning and development team had spent approximately two years trying to manually map skills across the organization. The effort stalled. The sheer diversity of roles — from customs compliance analysts in Frankfurt to parcel handlers in Lagos — made a traditional taxonomy project unworkable at DHL's scale.
"We needed something that could handle the complexity without taking another decade," said Meredith Wellard, Vice President of Group Learning, Talent and Platforms at DHL, describing the impetus for exploring AI-driven alternatives.
The core challenge was structural. Desk-based and field-based workers operate under radically different job architectures. A system that could infer skills for a supply-chain strategist also needed to meaningfully profile a driver whose expertise lives in route optimization, vehicle handling, and safety compliance — skills rarely captured in traditional HR systems.
The Solution: Cornerstone Galaxy AI and Its Clustree Origins
In 2020, DHL launched a pilot with Clustree, a French AI startup specializing in skills inference. Clustree's technology could analyze job descriptions, employee profiles, and organizational data to build a skills taxonomy automatically — no multi-year manual mapping required.
Cornerstone OnDemand subsequently acquired Clustree, folding its AI engine into what is now Cornerstone Galaxy — a platform that combines skills intelligence with career pathing and internal talent marketplace capabilities.
The results from the initial deployment were striking. The AI achieved 85% skills accuracy in its first off-the-shelf run, completing in less than five minutes what DHL's L&D team had been unable to finish in two years of manual effort. That number — 85% accuracy in under five minutes — became the internal proof point that shifted DHL from pilot to enterprise-wide rollout.
What the Platform Actually Does
The Cornerstone Galaxy-powered marketplace gives DHL employees visibility into internal career opportunities they would never have found through traditional channels. Employees can raise their hand for open roles, and recruiters gain line of sight into internal talent pools that extend well beyond their usual application pipelines.
For field workers, this is transformational. A warehouse operative with ten years of logistics experience and informal leadership skills can now surface as a candidate for supervisory or operations-planning roles — positions they might never have applied for, and that recruiters might never have considered them for.
One employee captured the shift plainly: "I have been at the company for 20 years, and this is the first time I felt that anyone has really bothered to think about me."
Measurable Results
By 2025, DHL reported measurable business outcomes from the skills marketplace deployment:
- More than 10% reduction in external recruitment resources, translating to millions in savings across DHL's global hiring operation
- 85% skills accuracy from the AI's first pass, validated against existing role profiles
- Unified skills taxonomy covering both desk-based and field-based roles across 220+ countries
The external recruitment reduction is particularly notable. In a labor market where logistics companies compete fiercely for hourly workers, reducing dependency on external hiring by even a single-digit percentage has an outsized financial impact. At DHL's scale, 10% translates into thousands of roles filled internally rather than through agencies, job boards, or headhunters.
What Other Large Employers Can Replicate
DHL's playbook offers three lessons for enterprises considering AI-powered skills marketplaces:
1. Start with field workers, not despite the complexity but because of it. Most companies deploy skills tech for knowledge workers first and promise to "extend it later." DHL's decision to include its 400,000 field workers from the outset forced the AI to handle real-world complexity — and proved the technology could manage it. If your skills platform cannot profile a driver or a warehouse worker, it does not cover your actual workforce.
2. Let the AI do the taxonomy work. DHL's two-year manual mapping effort was not a failure of effort — it was a failure of approach. Skills taxonomies at scale are exactly the kind of classification problem where AI outperforms human teams. The 85%-in-five-minutes result is not magic; it is what happens when you stop asking humans to do machine work.
3. Measure recruitment displacement, not just engagement. Many internal mobility platforms report adoption metrics — logins, profiles completed, courses started. DHL tracked what matters: how many external hires the marketplace displaced. A 10%+ reduction in external recruitment resources is a CFO-grade metric, not an HR vanity number.
The logistics giant's experience suggests that the barrier to enterprise skills transformation is not technology — it is the willingness to include the entire workforce from day one. For the 400,000 DHL employees who never sit at a desk, that decision made the difference between being an afterthought and being seen.
Sources:
How did DHL use AI to build its skills marketplace?
DHL partnered with Clustree (now part of Cornerstone Galaxy) to use AI to automatically infer and map skills across all 570,000 employees — including 400,000 field workers. The AI achieved 85% skills accuracy in under five minutes, compared to a two-year manual mapping effort that had stalled.
What results did DHL achieve from its AI skills marketplace?
DHL reported a more than 10% reduction in external recruitment resources, translating to millions in savings. The platform also created a unified skills taxonomy covering desk-based and field-based roles across 220+ countries.
What lessons can other large employers take from DHL's approach?
Three key lessons: include field workers from day one (not as an afterthought), let AI handle the taxonomy classification work rather than doing it manually, and measure recruitment displacement metrics rather than just platform adoption numbers.