TechWolf

TechWolf reads the work your people already produce, in systems such as Jira, Salesforce and Microsoft Teams, and writes the skills it finds back into Workday or SAP. Of the 139 skills products in our buyer’s guide, it is one of 18 that work out what people can do without being told. TechWolf does not publish a price.

In one line

Software for large companies that works out people’s skills from their work and writes them into Workday or SAP, suggesting how good somebody is only at the skills you agree are critical.

Buy it if

You run Workday or SAP, your people work in systems that hold a lot of text, and you have a change programme behind it rather than an integration budget.

Do not buy it if

Most of your workforce is frontline, which is where confirmation lagged at T-Mobile, or you need a proficiency level on every skill rather than only on the critical ones.

Price

Not published. You only get a number by going through TechWolf’s sales process.

Weighing it against Microsoft People Skills? Read the two side by side. The decision comes down to whether you need the skills worked out from people’s work to reach your HR system, which TechWolf does and Microsoft People Skills does not.

How it works

What it does

It sits underneath your HR system rather than giving your employees another system to log into. It connects to three groups of systems: your HR and recruiting software (Workday, SAP SuccessFactors, Oracle, iCIMS), the systems where the work happens (Jira, Confluence, GitHub, ServiceNow, Salesforce, Microsoft Teams, Slack, Google Drive), and your learning platform. It runs language models over the text held in those systems and works out the skills of each employee and each job, along with the tasks each job involves. It then writes the results back into Workday Skills Cloud, SAP SuccessFactors or Oracle HCM, and can feed them into reporting tools such as Power BI and Tableau.

TechWolf publishes some of its language models openly, so your own technical people can download them and try them on your data before you sign anything. They are not the whole system you would be buying, though: TechWolf had 15 language models in live use in 2024, and its public model library holds three, so a trial tells you about part of the product rather than all of it. It is also built to leave Workday or SAP as the system of record, so your employees keep the HR system they already use.

Working alongside those systems rather than replacing them keeps TechWolf close to the large software companies it connects to. SAP, Workday Ventures and ServiceNow Ventures all invested in its 2024 funding round, Workday and ServiceNow are also customers, and Workday Skills Cloud is what TechWolf writes into. Workday already suggests skills from information held in Workday and works some of them out itself, but it does not read systems such as Jira or Salesforce. Reading those systems is the part TechWolf adds, and it is the part you would no longer need if Workday ever built it.

What TechWolf gives you is a first draft of every employee’s skills, taken from their work and written into the HR system you already run, which your people then confirm and your business owners keep in order.

Its figures

What its figures measure

TechWolf’s published figures count different things, from accuracy to how many people accepted what it suggested, so read each one for what it counts.

about 95% accuracyTechWolf’s product pages

TechWolf puts its accuracy at about 95%, and it runs WorkRB, an open benchmark for this kind of software. One customer pilot across 94 job codes reported 99% skill accuracy. Ask what the headline figure measures, and on whose data, before it goes into a business case.

more than 77% of pilot users accepted the skills it suggested for themEricsson

The figure counts pilot users who agreed with the first draft of their skills, and the case study does not give how many of the suggested skills people turned down. A result like that supports using the suggestions as a first draft that somebody checks, rather than letting them feed a decision with nobody looking at them.

company-wide validation of around 32%T-Mobile

T-Mobile left it to employees to confirm their skills and never made it compulsory. Back-office functions reached about 70%, and most of the gap was in frontline retail and care, so the figure shows what happens when confirming skills is voluntary.

three to six months to validated skillsTime to value

Two to six weeks of that is technical set-up. The rest goes on checking the results and managing the change, which takes people from the business as well as time: Ericsson made more than twenty of them owners of its skill areas. At T-Mobile the checking has to keep pace with the roles themselves: its role architecture has grown from about 2,500 roles to nearly 4,000, with new roles created all the time, and each new role means more skills to check.

Taken together, the figures describe a strong first draft that a person still has to check.

Fit

Who it suits

Your people leave text trails

It reads what people write, so the more your people write, the fuller the picture it builds. That favours engineers working in Jira, consultants in Confluence, sales teams in Salesforce and anyone who writes documentation.

You run Workday or SAP and the skills fields are empty

This is the clearest job it does: it works out the skills and writes them into Workday Skills Cloud or SAP SuccessFactors, and whichever you run stays the system of record. Employees confirm the skills through an assistant in Teams or Slack, and in Workday they then see them on their own profile, so there is no new HR system for them to ignore.

A manual taxonomy project has already failed once

Genesys had spent 18 months mapping skills by hand for 700 job profiles, and the first profiles were out of date before the last ones were loaded into Workday. With TechWolf, it put together a skills pilot for 2,000 employees across most functions in under two weeks. Assembling a pilot group is a smaller task than mapping 700 job profiles, and the speed, set against 18 months of work by hand, is still the strongest part of the Genesys story.

You are in Europe and answer to a works council

At Belgium’s Federal Public Service for Health, the team took the project to the works councils early and showed them that the data sent to TechWolf was anonymised, which reassured them. TechWolf will not use LinkedIn profiles or any other external profile data, which in the EU means there is no outside data about your employees to explain to a works council.

Limits

Where it stops

A frontline or deskless workforce

At T-Mobile, when confirming skills was left to each person, the gap between the back office and the rest of the company was concentrated in frontline retail and care. If much of your workforce works away from a desk, plan how those people will confirm their skills before you rely on what the profiles say.

You need to know how good somebody is

Detecting that someone has a skill is easier than judging how good they are at it. TechWolf suggests a level only for the skills you agree are critical, and it does not run assessments. T-Mobile has a five-level proficiency scale but has not yet applied it to this work, and plans to grade proficiency only for the critical skills that change a hiring or deployment decision. Once people see skills they did not know they had, the next question at T-Mobile was how good those skills are, and if you plan succession or pay banding around skills, you will need that answer.

You want an outcome, not better data

It supplies the skills data, and the talent systems you already have keep making the decisions. At T-Mobile it is set to feed skills-based hiring in Workday HiredScore and internal moves in Opportunity Marketplace, rather than run either itself. HSBC found that knowing the gap is not the same as closing it, and the bank had to create a playbook setting out when to hire from outside, when to move people internally and when to retrain.

You are scoping this as an integration

The technical set-up takes two to six weeks and most of the calendar goes on checking results, setting up governance and managing change, so the budget needs to cover a change programme as well as the software. Expect to put a named person from the business in charge of each skill area, as Ericsson did with more than twenty.

Two smaller limits will shape what the profiles contain. It works out what a person can do only from data your own organisation holds, never from LinkedIn or other outside profiles, so a skill that leaves no trace in your own records will not show up. And if you bring an existing skill list, TechWolf typically maps 70 to 80% of it onto its own library straight away, and another 15 to 20% in the next round. The rest, usually in-house names such as internal tools, goes in as custom skills, which TechWolf does not yet suggest to anyone automatically. How the skills on your list are grouped stays yours to decide, although TechWolf says around 95% of its customers let it build and maintain their skills taxonomy.

Two industry analysts have written about TechWolf, Josh Bersin in 2024 and Brandon Hall Group in 2025 after a briefing from the company, and neither writes as a customer. On TrustRadius, a site where software buyers post reviews, its listing carries one rating and no published reviews, so ask how many customers are live today and speak to one close to your size. TechWolf holds SOC 2 Type 2 (an independent auditor’s report on its security controls), ISO 27001 (the international standard for information security) and ISO 42001 (the equivalent standard for managing AI), with the documents available on request. It reads performance reviews and HR records, so ask for the SOC 2 report and the certificates before you sign.
Money and size

What it costs

TechWolf does not publish a price. You will only get a number by going through its sales process, and it does not say in advance whether you pay per employee, per module, or a flat fee for the platform.

TechWolf states no minimum size either, so look at the size of the customers it names. Of the customers TechWolf gives a size for in its case studies, the smallest is Cotality, at about 5,000 people. Genesys comes next at around 7,000, and several are far larger: Ericsson, HSBC, T-Mobile and Atlas Copco each employ more than 50,000 people. Telenet’s logo also appears on TechWolf’s homepage, alongside T-Mobile and Genesys, and Telenet is smaller again, with 3,592 employees at the end of 2025 in its own annual report. TechWolf’s TrustRadius listing puts its audience at organisations with more than 1,000 employees, so ask where the line sits for a company your size.

There is a second route to this. The same job, built on the systems you already run, is usually a narrower piece of work, and we have mapped what that version looks like. See it as an agent

Before you sign

What to ask on the demo

  1. How reliable are the profiles for staff who spend little of their day in the systems you connect, and what share of them confirm their skills? Ask specifically about frontline and deskless roles, and ask for a comparable customer.
  2. Show me proficiency levels applied to a whole population rather than a sample.
  3. Which of our systems will you read to see the work people do, and will the tasks in each role come from that work or only from job descriptions? At T-Mobile, tasks are still worked out from job descriptions, not yet from the systems where the work happens.
  4. Where does the 95% accuracy figure come from, and how was it measured?
  5. Will you send us your SOC 2 Type 2 report and ISO 27001 certificate? TechWolf’s trust centre offers both on request.
  6. Ericsson has used TechWolf since 2022 and HSBC since 2023. Which customers have renewed, and can we speak to one?
  7. How many customers are live today, and how many of them are close to our size?
Other options

Alternatives to TechWolf, including buying nothing

The main alternatives to TechWolf are Microsoft People Skills and the skills modules in the HR suites you may already run: Workday Skills Cloud, SAP Talent Intelligence Hub, Oracle Dynamic Skills and Cornerstone. You can also buy nothing until you have seen what those already show you about your people.

If your people are on Microsoft 365, look at Microsoft People Skills first. It reads documents, emails, chats and meetings, and it costs $0 extra, so far, with Microsoft Copilot or Viva licences, so if you already hold one of those, it will show you how much your own systems know before you buy anything new. A business on Microsoft 365 alone gets People Skills, but not the part that works out people’s skills automatically. If you need to know how good people are as well as what they can do, the two differ: TechWolf’s proficiency scale has five levels by default, adjustable from three to eight, and Microsoft People Skills proficiency levels are not described in Microsoft’s People Skills documentation.

If you already own Workday, SAP, Oracle or Cornerstone, find out what their own skills features do before you pay for a separate product. All four keep skills on each person’s profile, not only on each role, and they set skill levels in different ways. Workday combines assessments with ratings from the person and their colleagues, SAP asks employees to rate themselves, Cornerstone blends a system prediction with self, manager and peer ratings, and Oracle can suggest a rating from a short assessment that the employee is free to overrule. Our overview of skills intelligence software sets out what you can do with a skills picture of each role, such as sizing a skills gap, and what needs a picture of each person, such as moving someone into a different role.

Check the licence before you count on any of it. SAP’s skills module comes with SuccessFactors, but working out skills automatically needs a separately purchased AI units licence. Oracle’s core HR subscription includes the skills module only “where applicable”, so ask Oracle whether yours includes it. Cornerstone includes a ready-made skills library with its learning and talent products and sells Skills Architect and People Graph as add-ons. Whether your Workday contract already includes Skills Cloud is a question to put to Workday.

Sources and method We read TechWolf's website, including its FAQ, published case studies and homepage, on 1 and 2 September 2026, and again on 14 September 2026. What TechWolf publishes about its own product is taken as its description of that product. We checked the language models TechWolf publishes openly, and the training data listed with them, on Hugging Face, the public website where they are posted. Funding and investor details come from TechWolf's own Series B press release. We confirmed that TechWolf's website has no pricing page. Customer headcounts are as stated in TechWolf's own case studies, apart from Telenet's, which comes from Telenet's own 2025 annual report. Security certifications come from TechWolf's trust centre. Analyst coverage comes from Josh Bersin and Brandon Hall Group. Details on Microsoft People Skills, Workday, SAP, Oracle and Cornerstone come from each company's own documentation, read on 14 September 2026.

On TrustRadius, the one review site we could open, TechWolf has one rating and no published reviews; G2 and Gartner Peer Insights could not be checked.