About Us
We are at the forefront of our field - and we want it to stay that way. Partly because it keeps us energized and sharp, but as importantly because it stands as an important promise of excellence to our customers.
Our Company
Alvenir traces its roots back to the Technical University of Denmark (DTU) from which the company is now a spin-out. While at DTU the founding team spent the better part of three years researching ways to improve automatic speech recognition in data scarce settings.
Funded by Innovation Fund Denmark the work at DTU led to several contributions including scientific publications, public talks and most importantly the first open-sourcing of an automatic speech recognition model for Danish.
Alvenir spun out from DTU in December 2021 and has since then continued to play an important role in the open-source community, staying true to its initial commitment, all the while continuing to expand its proprietary services and offerings.
Today Alvenir primarily operates in the financial industry, with its primary business focused on creating and operating specialized AI models to augment and enhance financial advisors in their dialogues.
Vision
Alvenir's vision is to revolutionize information capture and analysis using voice-based natural language processing. Our commitment to cutting-edge research and open-source development drives AI innovation for transformative advancements in our field.
Vision
Alvenir's vision is to revolutionize information capture and analysis using voice-based natural language processing. Our commitment to cutting-edge research and open-source development drives AI innovation for transformative advancements in our field.
Meet the Team
Meet The Team
Mission
Our mission is to develop and operate cutting-edge AI solutions empowering individuals and organizations to capture, interpret, and leverage valuable information from human interactions.
Alvenir In The Press
Tech Savyy
Alvenir: Research And Open Source Can Be Combined With a Commercial Startup
The startup Alvenir has built a scalable platform for speech recognition based on research from DTU. A platform that already speaks better Danish than Google and Amazon. The goal is a big business, yet the language models behind it are made available to others for free through open source.
Teknologiens Mediehus
Start-up company open-sources a pre-trained model for Danish speech recognition: Together we can make an even better version
Denmark has a new language model that can be used to develop speech technology. In December, start-up company Alvenir launched a wav2vec2 model that can be fine-tuned for Danish speech recognition and beats the international models by far.
Apple Podcasts
#36 Alvenir - Specialiseret dansk NLP
I denne episode kommer Martin Carsten Nielsen fra den danske start-up Alvenir i studiet. Alvenir er sprunget ud af et speciale på DTU, hvor de to studerende efter et års tid, slog sig sammen med den vejledende professor, og i dag angriber Fin-tech markedet, med specialiseret NLP. I stedet for at udvikle store generelle danske modeller, tilpasser de nemlig de eksisterende og udvikler dem, til Fin-tech. Vi taler også om Alvenir som platform og demokratisering af data.
Applied AI Lunch Talks 2022
How data-centric machine-learning can enable new business cases Alvenir
This lunch talk focused on real-life applications of speech and voice recognition in the financial sector. Focused on how to capture value with a variety of workflow applications and dived a little deeper into the hidden caveats of off-the-shelf speech-and-voice solutions before ending with a discussion on how specialization of the underlying machine-learning models can help overcome some of the most common issues.
DR.dk
Godaw’ do: Nu skal danske dialekter samles i en digital bank
Projektet skal sikre, at stemmestyret teknologi kan forstå de mange danske dialekter. Indsamlingen af dialekter kommer rundt i hele landet, og bliver skudt i gang i dag i Nordjylland. På sigt skal de mange timers tale blandt andet hjælpe medarbejdere på alarmcentralen med at forudse hjertestop.
Need help realizing your speech recognition project?
Don’t hesitate to send us an e-mail - we will get back to you as soon as possible.