Enterprise Doesn't Want AI to Screen Resumes – It Wants AI to Redeploy People

01 /The Concept

TechWolf "connects a company's business strategy with its people strategy."

Behind that tagline sits a technology that tracks the "digital traces" employees leave behind – emails, reports, presentations, task lists – and assembles a map of their actual skills.

Those traces are pulled automatically from integrated enterprise systems: documents in SharePoint, tickets in Asana and Jira, wikis in Confluence, conversations in Microsoft Teams.

The key insight is that job titles are almost meaningless. What matters is the role a person actually plays in workflows and the capabilities they demonstrate on real tasks. And those capabilities, in turn, are the visible expression of underlying skills that no org chart can see.

In other words: skills are fundamental. They surface as capabilities, which enable roles, which determine job assignments – not the other way around.

The platform delivers four concrete outcomes. Every type of work in the company – every project, every process step – is automatically tagged with the skills required to perform it. Every employee is simultaneously tagged with the skills they already have, derived from their digital activity. For each employee in a given role, the platform surfaces a gap list of skills not yet demonstrated and can automatically recommend learning content to close those gaps. And when a new position needs to be filled, companies don't have to go to market at all – TechWolf can surface internal candidates who are already qualified, or who need to learn just a handful of things to be ready.

The skill extraction engine is a purpose-built AI model trained on large volumes of professional data spanning many occupations and skill taxonomies. Because it's specialized, TechWolf claims it outperforms GPT-4 on this specific family of tasks by roughly 100x.

The platform is already in use at GSK, HSBC, Booking.com, Workday, United Airlines, Ericsson, and others.

Revenue has grown 12x in the two years since the previous funding round. Although TechWolf is a Belgian company, 50% of its existing and prospective customers are in the US – which is why the startup is opening a US office this fall.

TechWolf just closed $42.75M in new funding, bringing total raised to $55.6M.

02 /Why It Matters

Let's walk through the argument TechWolf used in its Series B pitch deck.

The most significant shift: large companies are changing their operating model. They used to run on org charts and formal job titles. Increasingly, they want to deploy specific people with specific skills – and the title on the badge matters less than what the person can actually do.

This is how startups and small companies have always worked, of course. But as organizations scale, that flexibility ossifies into formal hierarchy, where title outweighs capability.

Companies aren't opposed to returning to a skills-based model. The problem is execution: instead of a simple headcount plan, you need a complex map that connects tasks to required skills. And skills can't be invented ad hoc – that creates chaos. They need to come from a formalized taxonomy where every skill has a precise definition and a consistent name. TechWolf's AI builds that taxonomy automatically for each company, with a clean interface for manual overrides.

TechWolf is API-first, so it integrates with any internal or external system – HRIS, learning management, even labor market data showing which skills are in demand externally, what they pay, and where supply is tight.

The company defines its target customers as "mature" organizations, with a specific meaning. "Immature" companies solve HR problems like "fill this open position with this job title." Mature companies want to find someone who can solve a specific business problem – and would prefer to find that person internally.

For companies with several thousand employees, smart internal mobility can meaningfully reduce external hiring and retraining. At tens of thousands of employees, the savings become substantial.

This conclusion emerged from TechWolf's own pivot: they originally built AI for resume screening, but discovered that large enterprises would rather not hire externally at all if they can avoid it.

Gloat, [covered here](/review/vremja-ispolzovat-vnutrennie-rezervy) in 2022, is moving in the same direction – an internal talent marketplace for finding employees to promote or borrow for cross-functional projects. It has raised $192.6M. GoFIGR (formerly Flow of Work), covered in 2023, took a similar approach and raised $3.5M.

Another angle: Elqano, [covered here](/review/bolshim-kompanijam-jeto-nuzhno) in late 2022, also analyzes digital traces to map skills – but the use case is peer discovery: employees finding colleagues to consult on specific problems. They've raised €900K. Beatrust, [covered here](/review/kak-najti-togo-kto-nuzhen) in early 2022, tackles the same problem and has raised the equivalent of $9.1M.

03 /Opportunities

The core insight from TechWolf's story: external hiring is often a reflexive solution – the default move for companies that haven't thought harder about the problem. Large, mature organizations increasingly prefer to reconfigure the talent they already have.

That reframing opens an interesting question. How many other problems that companies routinely declare – and that startups rush to solve head-on – are actually solvable with a completely different approach? How many "we need to hire faster" problems are really "we need to find the right people internally" problems in disguise?

The broader lesson: take a company's stated problem, restate it, and propose a solution to the restated version – one that gets the company to the same outcome with less money, time, and friction.

Here's a rough illustration. Many startups say they can't hire enough qualified people to execute their roadmap. Maybe what they actually need isn't an HR tool – maybe they need a smart task prioritization tool to reduce how many people they need in the first place. Because "beat the competition through broader features" is not a startup strategy. Startups should be testing one breakthrough hypothesis at a time – the one requiring minimum resources but delivering maximum signal.

If problem-reframing isn't your thing, the more direct path is to build an equivalent skills intelligence platform. The appeal is obvious: the customers are large enterprises that pay meaningful per-seat fees at scale, and they're the kind of organizations that buy for ROI rather than novelty. That's a pleasant customer to have – if you have something real to offer them.