.avif)
Table of Contents
In recent years, Aliru GmbH – together with its AI solution Sally - has become one of the most innovative technology partners in the field of AI-powered meeting and sales processes. To accelerate this development and actively shape future technological standards, Aliru has been working since 2024 on a joint ZIM project within an official research collaboration with the University of Mannheim, one of Europe’s leading institutions in economics and data science.
Together, we aim to establish new foundations for high-precision speech recognition, meeting analysis and automated follow-up - especially for real-world meeting environments where existing systems reach their limits.
Sally already achieves up to 98.8% transcription accuracy, significantly surpassing the current industry standard. But for us, this is just the beginning. The collaboration with the University of Mannheim enables us to push the boundaries of speech recognition even further and address the challenges of real meetings with greater accuracy and reliability.
Why this ZIM-Project matters
Companies today use a wide range of tools to transcribe or summarize meetings. But as soon as situations become technically demanding - multiple speakers, shifting acoustics, background noise, jargon — quality issues arise. Many tools offer individual features, but not the level of precision and reliability required for business-critical conversations.
This is exactly where the joint research project between the University of Mannheim and Aliru comes in. The work focuses on:
- Improved speaker recognition in dynamic meeting situations
- Precise action-item detection for reliably identifying tasks
- Noise-resistant transcription that reflects real meeting rooms
- Recognition of industry-specific jargon rather than relying on generic language models
These capabilities already exist in partial form across the market - but not with the quality, robustness and integrated approach that companies truly need. This is why the research partnership between Aliru and the University of Mannheim is an important step toward elevating current technologies to a new level.
Rethinking Transcription: A Leap Beyond the Industry Standard
With up to 98.8% accuracy, Sally already delivers meeting transcripts that come closer to reality than nearly any other solution currently available. The joint ZIM project aims to extend this precision even further — particularly in challenging scenarios:
- conversations with multiple overlapping speakers
- meetings where participants move or change distance to the microphone
- environments with chair noise, typing, or background chatter
- company-specific terminology, product names, and abbreviations
The University of Mannheim contributes scientific depth and methodological expertise, while Aliru draws on an extensive set of realistic test data and many years of practical experience. These datasets were created specifically for training and development purposes and replicate typical meeting situations - without any connection to actual customer information.
Together, they provide a research foundation that is both scientifically robust and fully compliant with data protection standards.
Science + Practice = A Strong Signal for Our Users
The joint ZIM research project demonstrates that Sally is not a quick experiment or a typical off-the-shelf AI tool. Behind our technology stand scientific methods, publicly funded development work and one of Germany’s most respected universities.
For our users, this means one thing:
Sally is not just being improved - Sally is being advanced through scientific research.
Try meeting transcription now!
Experience how effortless meeting notes can be – try Sally free for 4 weeks.
Test NowOr: Arrange a Demo Appointment
.avif)
