Calls: Allow manual selection of source and target language for call transcript translation
R
Ricardo Gonzalez
Business Problem:
Call transcripts are auto-generated in the language spoken during the call, with no way to translate them or override the detected language. This creates three issues for multilingual teams:
- Cross-language review gap: Supervisors and managers who speak a different language than the customer cannot read or review transcripts in a language they understand.
- Regional dialect mismatch: Teams in Spain reviewing Latin American Spanish calls, or teams in Brazil reviewing European Portuguese calls, encounter translated content that doesn't reflect the correct regional terminology, tone, or vocabulary.
- Look-alike language misdetection: Closely related languages (Spanish/Portuguese, Malay/Indonesian, Hindi/Urdu, Serbian/Croatian) are frequently misidentified by automatic language detection, producing inaccurate or misleading transcripts that erode trust and require manual verification.
Desired Outcome:
Allow users to manually select both the source language (what was spoken during the call) and the target language for translation directly within the transcript viewer. This ensures transcripts are accurate, culturally appropriate, and usable across all regions and languages a business operates in.