Techwolf's Series B, and a deeper reflection on skills management and AI.

Applications are only as good as the data is truthful.

Today Techwolf announced its Series B funding of $42,75 million.

Why now and why is Techwolf different?

In HR tech circles, skills management is a big deal. There are 100s of vendors saying have skills powered this, and skills powered that. Esteemed consulting firms produce impressive reports with beautiful infographics about skills management and the holy grail of the skills based organization. Recent developments in AI have only added to the cacophony. It has become a category, but it is in ferment. Incumbent vendors in core HR, Learning, and Recruitment argue they have the answer, as do a newer wave of well funded players. They all promise a skills powered nirvana, but there is an excess of marketecture, rather than of actual delivery.

One of the three fundamental tenets of our investment approach is to ask the founders what do they believe that others don’t believe. While everyone else is out there building functionality, and there are great use cases in internal mobility, recruitment, and more, Techwolf is focused on the singular challenge of discovering, deriving, categorizing, pruning, inferring, scrubbing, polishing and curating skills data.

Getting great skills data into a useable shape is a wickedly difficult technical problem. Almost every vendor thinks is going to be easy. It is easy to do poorly.

Andreas, Mikael and Jeroen spent several years at university digging into the questions of how to derive skills data from unstructured data sources, gnarly issues of data providence, semantics, ontology management and more. They have deep geek chops.

For decades HR applications have suffered from functional excess and data poverty. The question where does the data come from in the first place receives scant attention, and it usually involves employees typing something in, or at best interfacing with other HR systems. it turns out that relying on employees to type stuff accurately and repeatedly without immediate payback isn’t a recipe for success.

Techwolf realised that the best skills data isn’t necessarily in the HR systems landscape, it is in operational systems and external data sources. While applying a responsible GDPR strategy from early on, it has now built the richest and most coherent skills dataset in the industry, with the most sophisticated AI capabilities to leverage that data, and an API first approach to sharing it.

Over the past five years, the company has published seven peer-reviewed articles and has 15 language models in production. Its datasets include 1.47 billion labour market data points, 38 million captured skill events, and 223 thousand recognised skill forms.

Scaling it

Over the last couple of years, Techwolf has proven out this thesis, winning deals and going live at some of the most complex organizations in the world. These organizations have deployed Workday, SuccessFactor, multiple recruitment and learning applications, and Techwolf now provides the skills infrastructure to make these applications work better. Techwolf is becoming a critical and fundamental component of the enterprise infrastructure. This has implications beyond traditional HR applications. This is, in a meaningful sense, multi-decade data infrastructure play.

For instance, GSK mines medical research data such as journal publications to derive granular skills data with Techwolf. This helps researchers discover peers and ultimately speed up research. This data then goes on to make Workday and a whole slew other applications more effective too, as it has more granular and accurate data for processes like internal mobility.

Techwolf has now built deep trust with both large enterprises, and several of the megavendors.

I met the team at Techwolf well before I moved in VC, and Acadian invested in the Series A, which was led by Fred (I learn a lot from him) at Stride. We are thrilled to continue investing with the Series B. Over the last few years, we have developed a deep collaboration with the founders and other investors, and look forward to more of this.

The decisions of ServiceNow, Workday and SAP to all invest in this round highlights the differentiated position Techwolf has established, and the strategic importance of what they do. The partner from the lead investor, Felix Capital is Julien Codorniou, who led Meta’s workplace. Semper Virens, a specialist VC in worktech is joining, so too Andy Leaver from Notion, who led both SuccessFactors and Workday sales in EMEA, As are 20VC, who need little introduction. Existing investors Fortino, PMV and Stride are also following on. Diane Gherson, the former CHRO of IBM is joining the board.

You may be wondering what the other two tenets of our investment criteria are. Simple. Do you really really really want to build a fund returning business? And are you a talent magnet? It is clear the founders of techwolf are determined to build a fund returning business, this was obvious from the first day I met them. They have already attracted top employees, customers and investors. If you are looking to be part of one of the most important companies in the industry, check out their jobs.

It is my tradition (when I remember) to end with a song. I couldn’t resist this LP cover from my collection, by the Manic Street Preachers. No prizes for guessing why I picked this one.

Have a listen here.