Textkernel matches CVs to job openings with language technology honed over more than twenty years

Anyone who has manually worked through a stack of CVs knows how time-consuming it is. Textkernel, an Amsterdam-based company founded in 2001 by AI researcher Jakub Zavrel, builds software that automates that process: from parsing a CV to proposing the best-fitting candidate for an open position. Over more than two decades it grew into a provider used by recruitment agencies and HR departments worldwide.

Since early 2024, Textkernel has been part of Bullhorn, a US software company focused on the recruitment industry. The acquisition was valued at approximately €300 million. The Amsterdam team continues to function as an independent unit within that whole, with its own product line and a focus that remains firmly on language technology for the labour market.

Four products for the recruitment process

Textkernel's offering rests on four core products, each targeting a distinct step in the recruitment process. Parser reads structured and unstructured text from CVs and job descriptions and converts it into usable, structured data. That may sound technically straightforward, but it requires complex language technology: a CV in Dutch is organised differently from one in German or Polish, and job titles vary considerably across sectors and countries.

Source & Match builds on that parsed data and helps recruiters search for and rank candidates based on their profile and an open vacancy. Skills Intelligence adds a taxonomic layer on top of that: the system recognises skills in texts and maps them to a standardised structure, so that a 'projectleider bouw' and a 'construction project manager' are recognised as related profiles. Finally, Labor Market Insights provides data on labour market trends, such as demand for specific skills or scarcity within particular job categories.

Together these products form a chain: raw text becomes data, data becomes insight, insight becomes a recommendation. Recruitment agencies and large employers integrate the software via an API into their own systems or ATS platforms.

Language technology as a foundation

What sets Textkernel apart from generic search systems is its emphasis on language understanding. Conventional keyword-based systems miss candidates the moment phrasing diverges from what appears in the job posting. Textkernel uses natural language processing to recognise the meaning behind words, not just their form. A system that understands that 'machine learning engineer' and 'ML-ontwikkelaar' refer to the same profession retrieves more relevant profiles.

This technological approach is rooted in the background of founder Zavrel. He is by training an AI and NLP researcher, and founded Textkernel with the ambition of building independent, technology-driven companies for the long term. That focus on language technology remained the central guiding principle over the years, even as the company grew and expanded its product portfolio.

The software is offered in multiple languages and is tailored to the labour markets of different countries, with local differences in CV formatting, job titles and sector terminology incorporated into the models.

From Amsterdam startup to part of an international platform

Over more than twenty years, Textkernel followed a trajectory that will be familiar to many Dutch tech companies: built in Amsterdam, grown internationally, and ultimately acquired by a foreign party. The acquisition by Bullhorn, completed in January 2024, fits within a broader movement in which specialised European HR-tech companies are being integrated into larger platforms seeking to cover the full recruitment cycle.

For Textkernel, joining Bullhorn means access to a larger customer network and greater resources for product development. The Amsterdam office remains the centre of gravity for technical development. Zavrel himself is now working on a new AI-based platform, though the details have not yet been made public.

For the Dutch and European HR-tech sector, Textkernel illustrates how a niche player in language technology can operate independently for years on the basis of deep domain expertise. The fact that such companies ultimately become attractive to large international parties says something about the value of specialised AI expertise in a market that increasingly wants to operate at scale. For founders and investors in the European HR-tech space, Textkernel thus serves as an example of a long-term trajectory: not a quick exit, but a gradual build-up of technological depth that ultimately carries its own weight.