English Speech Recognition Technology
powered by Global ESL student data
To provide the most effective English learning experience,
Databank’s Speech-to-Text (STT) model is trained solely on data collected from non-native English speakers around the world.
STT technology optimized
for ESL speakers
Speech recognition specialized for ESL speakers
Databank’s STT model is trained with over 12,000 hours of speech data collected from more than 1.4 million ESL speakers.
Client specific needs and customization can be met using techniques including context based recognition and transfer learning.
Speech assessment solution
Databank’s STT technology computes user specific recognition confidence metrics which can be utilized to evaluate speaker’s skill level and provide user specific content recommendation.
Databank’s efficient STT architecture design allows for cost savings of up-to 50% compared to other similar services.
[Real case example] Comparison of the STT costs of Company A/G and Databank (22.05.2021 - 21.06.2021)