HumanSignal Launches Multimodal Data Services with Acquisition of Erud AI
New Services division will deliver novel, high-quality data creation for frontier AI labs and enterprise model builders
SAN FRANCISCO, Nov. 05, 2025 (GLOBE NEWSWIRE) -- HumanSignal, creator of the industry’s most flexible open source data labeling platform, Label Studio, today announced the launch of HumanSignal Services, a new division focused on multimodal data creation and expert annotation for the next generation of AI systems.
As AI progresses beyond text and images toward embodied, multimodal intelligence, the industry faces a shortage of untapped, high-quality data. HumanSignal Services addresses this challenge by creating novel, real-world datasets—from on-site data collection to controlled multimodal generation for audio, vision, robotics, and sensor-based models.
“AI has outgrown the internet,” said Michael Malyuk, Co-Founder and CEO of HumanSignal. “The next frontier of intelligence won’t come from re-labeling or refining what already exists. It will depend on purpose-built, high-signal data created with human expertise, security, and precision. That’s what HumanSignal Services was built to deliver.”
The launch follows HumanSignal’s acquisition of Erud AI, a leader in physical data collection and domain-specific data creation. Erud’s team brings deep operational expertise scaling secure, on-site data facilities and workforces. Together, HumanSignal and Erud will power an end-to-end data creation factory, combining Label Studio’s multimodal platform and quality control workflows with operational services.
“Our customers need more than a remote workforce marketplace,” said Alexandra McGauley, Co-Founder of Erud AI and now President of Services at HumanSignal. “We partner with them from concept to delivery—designing, creating, and validating the data their systems need to learn from the real world.”
Initial Offerings
- HumanSignal Services will focus on high-impact, operationally complex data creation for frontier AI labs, robotics, and enterprise model builders, including:
- On-site multimodal data collection — capturing human interactions, sensor data, biometrics, and real-world scenarios.
- Robotics and embodied AI — synchronized motion, vision, and audio data from controlled physical environments.
- Next-generation device data — audio, sensor, and visual data collection from prototype and unreleased hardware.
- Expert annotation and evaluation — high-touch recruiting, training, and management of expert AI trainers for labeling, evaluation, and red teaming.
Extending the Power of Label Studio
Used by hundreds of leading AI companies and more than 500,000 open source users, Label Studio remains the foundation for human-in-the-loop data creation and evaluation. Its enterprise version delivers unmatched flexibility, auditability, and quality control for complex, multimodal workflows.
With the launch of HumanSignal Services, the company now offers a complete data stack—combining software, infrastructure, and operational expertise to help enterprises embed proprietary knowledge into models and enable AI labs to build the datasets no one else can.
For an inside look at HumanSignal Services, Malyuk and President of Services, Alexandra McGauley, take us to a frontier data lab to show how the team creates multimodal data from scratch.
About HumanSignal
HumanSignal is the company behind Label Studio, the most widely adopted open source data labeling and evaluation platform for AI development. Used by leading enterprises, research organizations, and AI labs worldwide, Label Studio powers flexible, multimodal data workflows for human-in-the-loop AI. With HumanSignal Services, the company now provides secure, end-to-end data creation operations to accelerate the next generation of AI.
CONTACT
Lauren Sells
lauren@humansignal.com
CONTACT Lauren Sells
Legal Disclaimer:
EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
