SAP acquires TechWolf to ground HR agents in real work data

SAP announced on Tuesday, October 6, 2026, that it has agreed to acquire TechWolf, a Belgian artificial intelligence company specializing in mapping employee skills and actual work activities. The deal was revealed at the Connect conference in Las Vegas and is expected to close in the fourth quarter of 2026, pending regulatory approval. This move aims to provide better context for SAP’s HR agents, specifically within the SuccessFactors portfolio, by grounding them in real-time data about what employees are actually doing rather than relying on static job descriptions or self-reported assessments.

What happened

TechWolf will continue to operate as an independent entity under its current leadership, with co-founder Andreas De Neve remaining as CEO. The company has previously maintained a long-standing partnership with SAP, which was among the investors in TechWolf’s $42.75 million Series B funding round in 2024. That round also included competitors like ServiceNow and Workday, highlighting the broad industry interest in TechWolf’s approach to skills inference. Existing joint customers have already seen benefits from integrating TechWolf’s inferred skills data into SuccessFactors modules such as the Talent Intelligence Hub, Recruiting, and Learning.

The primary technical goal of this acquisition is to enhance the performance and efficiency of SAP’s AI agents, known as Joule. By feeding these agents a proprietary "context graph" that details the specific tasks, skills, and labor market trends relevant to each employee, SAP intends to make agent queries more accurate. Manoj Swaminathan, president and chief product officer for SAP Autonomous Suite, stated that this graph "makes token usage more efficient" which helps cut the operational cost of running these HR agents. This efficiency gain mirrors similar claims made by other tech giants, such as Atlassian, regarding their own context graphs.

Key details

• SAP expects the acquisition to close in Q4 2026, subject to regulatory approval.
• TechWolf will remain an independent brand with Andreas De Neve staying on as CEO.
• The integration focuses on providing a "context graph" for skills and work to ground AI agent queries.
• TechWolf’s technology infers skills from actual work data rather than self-assessments or static resumes.
• The deal aims to reduce token usage costs for SAP’s Joule AI agents by improving query precision.
• TechWolf publishes open models like JobBERT on Hugging Face, supporting English, Spanish, German, and Chinese.

Background

To understand the significance of this deal, it helps to understand how large language models (LLMs) and AI agents function in enterprise settings. These models generate responses based on patterns in data, but they can hallucinate or provide generic answers if they lack specific context about a company’s internal operations. A "skills graph" or "context graph" acts as a structured database that maps relationships between employees, their daily tasks, the skills required for those tasks, and broader business goals. Instead of guessing what an employee knows, the AI can look up verified data derived from their actual work output.

Token efficiency is another critical concept here. In AI processing, a "token" is a unit of text that the model processes. The more context you feed an AI to ensure it gives a correct answer, the more tokens you consume, which directly increases computational costs. By providing a highly relevant, pre-structured graph of employee skills and work history, the AI needs less trial-and-error prompting to find the right candidate or suggest the right training. This reduces the number of tokens used per query, lowering the overall cost of running AI services at scale.

Why it matters

For teams managing their own software infrastructure or evaluating HR technologies, this acquisition highlights a shift toward "grounded" AI. Generic AI tools often struggle with internal talent management because they do not know the nuanced reality of who does what in your organization. By acquiring a tool that maps actual work to skills, SAP is addressing the core problem of enterprise AI: context. For IT leaders and HR managers, this signals that future HR platforms will need to integrate deeply with workflow tools to capture real-time activity data, rather than relying solely on periodic reviews or manual profile updates.

The focus on token efficiency also has practical financial implications. As companies deploy more AI agents for tasks like skills-based hiring, role redesign, and workforce planning, the volume of queries can skyrocket. Without efficient context layers, these costs can become prohibitive. Solutions that reduce token usage while maintaining accuracy will likely become standard requirements for any self-hosted or purchased HR suite. This deal reinforces the idea that AI value is not just in the model itself, but in the quality and structure of the data fed into it.

What you can do

• Audit your current HR data sources to see if they reflect actual work output or just static job titles.
• Evaluate whether your existing HR tools offer API access to integrate with workflow or project management systems.
• Consider the total cost of ownership for AI features, including potential token usage fees, when selecting vendors.
• Look for HR platforms that support dynamic skills mapping rather than relying solely on manual employee self-assessments.
• If you self-host HR solutions, explore open-source models for skills matching that can be integrated into your local stack.
• Prepare for increased demand for data privacy controls as AI agents gain deeper access to individual work patterns.