Knowledge Graphs are destined to power the future of talent management

Why you must use graph technology to solve the biggest challenges facing the HR industry…

Photo by Timon Studler on Unsplash

At Beamery, we use semantic web technology to build knowledge graphs representing organisations and their people. This means capturing data on everything from the candidate experience, through to off-boarding.

To help us explain this lifecycle I’ve borrowed some information from the Academy to Innovate HR (AIHR). They’ve done a great job at condensing this process into seven distinct steps:

With a little more detail and context on the underlying enabling technology:

1. Attraction — marketing and advertising your employer brand as a desirable place to work. Communicating what it is like on the ground.
2. Recruitment — sourcing, assessing and vetting candidates to join an organisation.
   Software Stack — applicant tracking system (ATS), Candidate Relationship Manager (CRM), referral platform, video interview software, identification verification software
3. Onboarding — equipping a new employee to succeed at their role.
   Software Stack — learning management system (LMS), human capital management system (HCM)
4. Retention — ensuring employees are effective and enjoying a role, and therefore likely to stay at an organisation.
   Software Stack — human capital management system (HCM)
5. Development — growing and progressing staff is a key part of retention.
   Software Stack — learning management system (LMS), internal mobility system, human capital management system (HCM)
6. Off-boarding — managing the process of an employee leaving in a professional and compliant manner.
   Software Stack — human capital management system (HCM), alumni management system
7. Happy Leavers — engaging previous employees. This can also involve re-targeting them as potential recruits.
   Software Stack — marketing automation system, content management system (CMS), alumni management system

Now, that’s a really large amount of software sprawl required to power such a complex operation. Now imagine you have four different legal entities from various mergers and acquisitions — you might now have 4 different software products for each system listed above. Just thinking about that amount of siloed data and complexity makes me nervous.

Software sprawl is the enemy of efficiency and productivity. Invariably many of the tools listed above will be underutilised as they overlap with one another, or don’t quite fit the needs of users. What’s more, sprawling systems can make compliance more complicated, bewilder users, and is very often more expensive than a 360 solution.

Imagine a world where your HR software stack ended up looking like this:

Everything does a bit of the lifecycle, but never all of it.

How have we ended up here?

For a long time, this status quo of various disparate HR systems has remained. In my opinion, there are two basic reasons for this:

• Humans are difficult to model at a point in time. We aren’t like machines — we don’t fit into clearly defined categories, have clear identifying numbers, nor do we have timestamps reported in a specific standard. It’s difficult to unify data with differing taxonomies and standards of reporting.
• Humans are difficult to model over time. There hasn’t ever been a single data source to track a person’s career over time. It just doesn’t exist. Instead, we are piecing together snippets.

There are various reasons that staying with this status quo is no-longer acceptable. Companies like Beamery are responding to this market need. Consider the following reasons for disruption in software servicing the talent lifecycle:

• Covid-19 pandemic driving talent to often migrate, and/or reevaluate occupations.
• Broader geopolitical trends to shun the benefits of globalisation, foreign talent acquisition, and introspective market conditions. This has tightened labour markets
• Technology advances are making skills more niche and harder to source.
• Record levels of employment driven by a long period of low interest rates, increasing labour market flexibility, and Brexit.
• Increasing numbers of highly technical roles have little or no formal vocational training pathways. For instance, software engineers lack a prescribed pathway via formal university education.
• Acceptance that diversity and inclusion is both a moral requirement, but also a business need.

Enter the knowledge graph

The venn diagram above is a nightmare for data management. Here at Beamery we are working to solve exactly that problem.

Now, if you don’t yet know the basic principles behind knowledge graphs, I have always used the following definition:

“A highly flexible no-sql database which represents data as “knowledge” through a graph-like structure of nodes and edges. Information is represented much like someone might draw a mindmap, or creatively related ideas together on a piece of paper. The nodes that refer to the knowledge are often defined in an ontology — the concepts that describe the domain. They can be traversed semantically using domain knowledge”

It’s because of this flexibility, and strong reliance on a strongly typed ontology that we are able to easily aggregate different data sources. We can represent data from multiple sources are closely related knowledge. Think of it a bit like this:

What we get in the knowledge graph is a unified view of all the original data sources. They are interlinked, normalised and up to date. This gives us the ultimate data source for talent intelligence, strategy and management.

For more information on the technical implementation of this, check out this previous blog I wrote. It’s a complex field, and not a trivial process.

What’s next?

Well, we’re continuing to build it. Stayed tuned.

If you enjoyed reading this article, follow our Beamery blog on our careers site or check out our open roles and apply!