### Answer --- Transcription accuracy for Indian accents has been a known challenge. Here are some approaches you can try: 1. Use Boosted Keywords Add domain-specific terms or commonly misheard words to enhance the model's vocabulary—you can add up to 100 custom keywords. [(1) ]() 2. Switch Transcription Mode to "Optimize for Accuracy" This mode uses higher endpointing settings, waiting longer with more context to generate more accurate transcripts (adds ~200ms latency) . This is especially helpful for capturing entities like names, numbers, and dates. [(2) ]() 3. Adjust Denoising Settings Play with the [denoising mode setting]() to see if it helps reduce errors caused by background noise or speech. [(1) ]() Recent ASR Improvements Retell's upgraded ASR engine has shown significant improvements for Hindi (WER improvements of 15–25 points) , with better handling of accent diversity and code-mixed speech. [(3) ]() Community Feedback Others have reported similar issues with Indian English accents. The Retell team has acknowledged this and added it to their backlog for improvement. [(4) ]() If these steps don't resolve the issue, consider sharing specific call IDs with Retell support for further investigation.

​ To learn more, see the sources I considered: [[1] Increase transcription accuracy](https://docs.retellai.com/reliability/wrong-transcript) [[2] Balance between transcription accuracy and latency](https://docs.retellai.com/build/transcription-mode) [[3] Smarter Understanding, Clearer Calls: Retell’s ASR Just Got a Major Upgrade | Retell AI](https://www.retellai.com/blog/retell-asr) [[4] Speech to Text for Indian English accent is very poor](https://discord.com/channels/1186424376300679209/1227468465330458684) ​

Feel free to tag @AI Evy with additional questions. I also appreciate feedback, it helps me improve.