speech-to-text services

Speech recognition

Our speech recognition services are ‘back-end’ products meaning audio is converted without the need for ‘there and then’ correction or reformatting. 
Audio is converted from speech into text using a bespoke, profile based version of Microsoft Azure’s

speech-to-text software and sent back to you as a completed document.  Users do not need to adopt a different process to begin their journey with artificial intelligence.


Rough draft (speech recognition)

Ideal for users needing a rough draft. Users do not sit and correct the document, wasting valuable professional time. The process of dictation remains unaltered; audio is run through the speech recognition engine and a block-text document is returned.


Proofread (speech recognition)

Once created by the speech recognition engine, documents are professionally proofread, eliminating software errors and formatting. This product is for users requiring a more complete document.

Okay for speech recognition:

Not great for speech recognition (use transcription):

How it works

Examples of speech recognition

Frequently asked questions

Still not found what you’re looking for?  Below is a list of FAQs that should help. 

Accuro has been transcribing confidential recordings for over 20 years.  Our software has been developed around the stringent encryption standards required by the NHS.  We are proud to be certified to ISO27001, the internal standard governing data security.  We are visited annually by external auditors who review every part of our transcription, operational and technical processes. We are registered with the ICO and have an inhouse compliance team who are responsible for monitoring adherence to UK GDPR regulations. 

We pride ourselves on the accuracy of the transcripts we return.  Human transcription will always be the best option, with your dedicated team being able to get used to accents, styles, interruptions etc. If however you have a budget restriction we have two other solutions that may be more useful.  Our AI solution with edit included will produce high levels of accuracy without the formatting that transcription provides.  Speech recognition alone will produce a lower level of accuracy but can still be a very useful tool if you need to quickly produce a rough draft.

Yes.  The service operates on an on-demand basis meaning you can dip in and out as often or infrequently as needed.  We do however offer discounts for large volumes or frequent uploads so it’s always worth speaking to your account manager about this.

We’d love to hear from you if you have experience in medical transcription, legal transcription or professional services transcription. Our recruitment team can guide you through the process and help you grow your panel of customers based on the areas you specialise in. To apply now, please click here.