Customer Spotlight

Deepgram is "Incredible" - Accuracy Up to 95%

Published: July 22, 2021
Customer Profile
Internet Software & Services

This Customer Spotlight is a synopsis of how a Small Business Internet Software & Services company used Deepgram to benefit their business. The feedback included was collected and verified in a January 2022 survey of Deepgram customers conducted by UserEvidence, an independent research firm.

Key Results


Accuracy Levels


Time to Go Live


Would Recommend

Overall, we've seen an average transcription accuracy of 90%. We've had some audios where the confidence is from 80 to 85%, but when the client's audios are high quality, the confidence can go up to 95%. It is incredible, because there are some potential clients who have not seen something of the kind in spanish, so this represents the highest impact metric we can talk about. It is not something you see everyday. Besides that, the diarization that is produced is really good, though we really hope it can improve. This has helped us move our demos quicker than before, so that we get in the client's pain sooner. We don't have real clients yet, but with our new personalized trained model, we believe we'll land them soon enough to start seeing our return.

Chief Artificial Intelligence Officer

Small Business Internet Software & Services Company


What were the main challenges that you were encountering that led you to purchase Deepgram?

  • Low accuracy from incumbent or alternative speech recognition engines
  • Inefficient on-premises hardware utilization; i.e. uses too much computing resources inefficient
  • Couldn’t transition to real-time use cases with current provider
  • High speech recognition usage costs that wouldn’t scale
  • Poor or nonexistent customer service

What competitors did you evaluate before choosing Deepgram?

  • Google STT
  • IBM Watson
  • Open Source
  • Nuance

What factors led you to select Deepgram?

  • Ability to increase accuracy with model training
  • Competitive Pricing
  • Deployment Flexibility
  • Customer Success and Ability to Partner
  • Out of the box accuracy
  • Overall Scalability

What accuracy levels were you seeing before Deepgram

  • 70-80%

Please select the options that describe your deployment

  • Trained model
  • Hosted by Deepgram

How long did it take you to go live in production once you decided to implement Deepgram?

  • 5-8 weeks

What accuracy levels are you seeing with Deepgram?

  • 85-90%

What speech recognition benefits are you achieving from your Deepgram deployment?

  • Improved Transcription Accuracy
  • Ability to differentiate from competitors

What business benefits are you seeing from choosing Deepgram?

  • Able to successfully launch a new product

What do you anticipate will be your timeframe for return on your investment (ROI) in Deepgram?

  • 6-12months
About Deepgram

Deepgram reinvented Automatic Speech Recognition (ASR) with a complete, deep learning model that allows companies to get faster, more accurate transcription, resulting in more reliable data sets — on-prem or in the cloud. And we do it all with lower hardware and usage costs so we’re a hell of a lot more scalable than big tech players.

By using this site, you agree to our Privacy Policy.