Last week, Daimler CEO Dieter Zetsche was at the White House for trade talks with President Trump, along with his peers from rival German automotive companies BMW and Volkswagen.
It will be one of the last high-level meetings that 65-year old Zetsche attends as CEO, before he hands over the reins of the company to current head of R&D Ola Kaellenius in May next year.
A particular feature of Zetsche’s final years as CEO has been his push to create highly entrepreneurial ‘swarms’ to speed up decision-making and empower staff at Daimler, a company more generally known for its strict hierarchies. A swarm is typically an interdisciplinary team, assembled temporarily to carry out a particular project and then disbanded.
The idea here is to remove layers of bureaucracy and encourage more experimental approaches in areas like electric vehicles and autonomous driving, at a time when Zetsche readily admits that he sees Google, Apple and Tesla as a bigger threat to Daimler’s Mercedes-Benz brand than traditional rivals Audi and BMW.
But the shift to being a more swarm-based organisation comes with some big demands when it comes to the visibility of human resources (HR) data, according to Jochen Linkohr, manager for HR IT at Daimler with responsibility for data integration, analytics and innovation.
In particular, when employees are constantly grouping and regrouping in swarms, it can be difficult to keep track of corporate structures, in terms of who employees report to - and who reports to them. It’s perfectly possible for an employee who works as part of one or more swarms to have more than one boss - but somehow, HR must still to be able to quickly find out about their current involvement in a project team, their workload or their immediate supervisor, for example.
This was the driver behind the deployment of a graph database at Daimler from Neo4j. Graph databases lend themselves well to the complexity of HR data, says Linkohr, whereas the neat row-and-column format of a relational database struggles to encompass the complex web of interconnections that exist in a workforce in terms of employees, departments, locations, training and skills.
Linkohr and his team used Neo4j as the basis of a new application at Daimler called StructureCube, which enables HR managers to see the multiple connections that can exist between individual employees, their reporting lines and their team affiliations.
This is a big help to HR staff, he says, in terms of maintaining oversight of corporate structures when it comes to personnel, even in the midst of regular changes. Plus, it gives them a way to uncover previously unknown connections between employees. As Linklohr explains it:
Someone working in HR can use StructureCube to see, for each employee, their supervisors, division, location, job function, colleagues and fellow team members. When the company restructures - for example, a new swarm is formed - the nodes representing employees can be moved about, but the integrity of the underlying data model is retained.
You might also see that ‘John’ is an expert on analytics, and ‘Amy’ is an HR expert and yes, they’ve worked together in the past on an HR analytics projects, so they would be good candidates for similar projects in future.
Daimler’s StructureCube app went live in late 2017, he says, and got a great reception from HR. Since then, he’s found new uses for the Neo4j graph database. In June this year, for example, he and his team introduced a new Committee tool, which enables HR to understand better how Daimler’s complex network of committees fits together. In this tool, committees can be searched, edited and approved, he explains.
So it’s possible to see which people are in a particular committee, what other committees they may be on, and their roles on each committee. And again, in a company like ours, there’s a lot of complexity here. You have committees with direct links to other committees, you have hierarchies of committees, you have subcommittees - and again, this is a great reason to use a graph database.
Now, he’s starting to think about a third tool - a recommendation engine that HR could use to identify those employees that have the right skills and experience to work on specific projects underway at the company. This is still in its early stages, he says, with his team just ‘playing around’ with ideas right now:
But come back in a year’s time and I hope we will be able to show that we’ve found yet another great use for graph databases in HR.