Researchers from the Ludwig Maximilian University of Munich (LMU) were working on a project for Horizon 2020, funded by the EU, to identify and build programmatic responses to expressed needs of vulnerable populations during the Covid-19 pandemic.
The research project consists of 5 phases over 1.5 years:
1. To develop, implement and analyze vulnerability assessments for Covid-19
2. To apply findings to actionable recommendations
3. To scale definitions of vulnerability to national health data set
4. To develop ongoing, integrated assessment training programs
5. To scale the program to global networks
The research team needed a partner to help them accurately transcribe their qualitative interviews into multiple languages in a safe and secure environment.
The problem:
During Phase 1, the ten-member research team held 83 2 hour-long interviews with participants from vulnerable populations in various languages, including German, English, Farsi, and Turkish.
To interpret the results of the interviews, they would first need transcribing.
“As you can imagine, it wasn’t an easy task. We had over 166 hours of content in many languages and only ten people on our team. ” – The research team.
With a long process ahead of them, 1 hour of interview recordings can take up to 5 hours of manual transcription.
As transcribing an -hour-long recording can take up to 5 hours of manual work, that would total 830 hours of manual transcription for the team. Divided between the 10 group members would mean 83h of transcription work for each member! The arduous process would mean stalling the project and taking the focus away from other areas.
“We needed a transcription provider that the whole team could use, fit in our tight budget, facilitate the languages we needed and handle sensitive data. Amberscript was the right solution.” – The research team.
AmberScript’s solution
Thanks to Amberscript’s automatic transcription tool and the Multi-User-Account, the team reduced the initially planned time spent on transcription of 830 hours by over 50%.
The results were achieved by the power of AI joining hands with a human touch in the form of an intuitive editor so that each researcher can improve the subtitle themselves for 100% accuracy.
“Amberscript offers a great variety of transcription languages – especially Farsi and Turkish were languages that other solutions didn’t cover. The account manager was also very helpful when someone on the team had a question with the platform. ” – The research team.
Amberscript’s solution results in much higher productivity and time savings and lower costs. With a world-class team of experts in technology, language, and science, Amberscript aims to make all audio accessible to positively impact society.