Reviews from AWS Marketplace
0 AWS reviews
-
5 star0
-
4 star0
-
3 star0
-
2 star0
-
1 star0
External reviews
External reviews are not included in the AWS star rating for the product.
Working with Speech-to-text made easy using AI
What do you like best about the product?
Documentation is the best because it clearly explains what each endpoint is supposed to do along with all the different parameters we can pass to tweak the results to our requirements.
What do you dislike about the product?
Support for more SDKs is required because Deepgram is a wonderful software but developers might be reluctant to using it if they are not able to find the SDK for their tech stack.
What problems is the product solving and how is that benefiting you?
I made a project where I needed to extract the first word in the spoken sentence depending on the context of the sentence. That way it solves the problem of homophones.
- Leave a Comment |
- Mark review as helpful
Excellence of Deepgram
What do you like best about the product?
I implemented deepgram ASR for my company's customer support application. I tested with their API and Even though I have no proper western accent still Deepgram provides their excellence. Processing time is also good. I faced a really a good experience with their speech recognition API call.
What do you dislike about the product?
For the past 8 months, I only faced one library issue which is related to an SSL certificate and even that also they supported my mail and then resolved it. The only thing is when pronouncing numbers/ some IDs with numbers for Audio recognition it is converting into words! Not numbers. It might not be a big issue but in a very rare scenario, it is not relevant data.
What problems is the product solving and how is that benefiting you?
I'm using Deepgram ASR for customer support and because of this my customers no need to type their text every time. I showed my previous implementations and projects to my company's new client and they were impressed. At first time tested deepgram ASR (Automatic Speech Recognition) using passage content in internet and yes the accuracy level was above 95%.
Recommendations to others considering the product:
If you are looking for ASR(Automatic Speech Recognition) specifically to customers/consumer support and tech support means deepgram is the best choice.
Good ASR performance but kinda slow model.
What do you like best about the product?
I liked the accuracy of the model and the built-in features for model prediction API.
What do you dislike about the product?
Slow model speed and running multiple jobs parallelly keep getting disconnected.
What problems is the product solving and how is that benefiting you?
Recognizing speech transcripts for many audio files.
Deepgram transcription save the day.
What do you like best about the product?
The support is excellent, and helpful, and having a direct line with the team that improves our models makes a difference
What do you dislike about the product?
if we could get a faster model life cycle it could improve but don't take this the wrong way, good models take time.
What problems is the product solving and how is that benefiting you?
We solve our Real-time and post-call transcription needs with an on-prem solution. The custom models help us understand better our type of calls and everything is handled in real-time.
Excellent product, good transcription, but needs usability features
What do you like best about the product?
The transcription capabilities are excellent, and the different models by use cases (video call, phone call, etc.) are fantastic. It's relatively easy to use, and the authentication mechanism is quite straightforward. I love how easy it is to set it up to begin with, and how quickly I coudl get started with prototyping, and not worry about "everything else" - it made getting my project off the ground incredibly easy!
What do you dislike about the product?
I wish there was more documentation around the different parameters that would let me tailor the API to different use cases. For example, in my use case, I need to proxy a file stream from third-party data sources to maximize client compatibility. I know the engineering to do it, but the SDK for Python does not support arbitrary buffer streams, and the API online is tailored to the SDKs instead of describing how to submit a manual POST request to post packets/a file stream. The lack of documentation just elongates development time. I can't use the streaming API because the transcription quality will not be quite as high as it would be in the case of a prerecorded audio file.
Another feature would be improved authentication mechanisms; while an API key is great, the ability to more securely encrypt transmissions, pointing Deepgram to an S3 bucket/blob would greatly improve the ability to integrate Deepgram for Enterprise clients.
Another feature would be improved authentication mechanisms; while an API key is great, the ability to more securely encrypt transmissions, pointing Deepgram to an S3 bucket/blob would greatly improve the ability to integrate Deepgram for Enterprise clients.
What problems is the product solving and how is that benefiting you?
I'm still prototyping, so I cannot point to success stories through this quite yet. I'm trying to improve VoC for SMB clients, but my alpha test customer wants to look at sales data analytics for competitor research. They have their audio files on a proprietary store; while they are comfortable sharing the audio with me, they need to be able to trust that the data is secure, so I am avoiding the need to store their audio files directly by proxying the file stream call (there is no way for Deepgram to authenticate with their source systems directly).
Recommendations to others considering the product:
Use this for an easy-to-roll-out prototype transcription service, but be ready for additional engineering to meet enterprise security parameters.
Deepgram renewal
What do you like best about the product?
The product works well is consistently being updated and improved. It is also the most affordable solution we have found.
What do you dislike about the product?
I don't have any complaints about Deepgram.
What problems is the product solving and how is that benefiting you?
We are using Deepgram for Transcription services and have found that the transcription of verbiage has helped us in translating audio to text.
excellent service - best price and highest quality transcription service.
What do you like best about the product?
Deepgram has the highest quality transcriptions. Additionally, they have the best pricing.
What do you dislike about the product?
With the release of their new console. they're the best transcription service.
What problems is the product solving and how is that benefiting you?
We're able to use Deepgram to provide high-quality call analysis in real time to customers.
Recommendations to others considering the product:
We used IBM Watson, Google, AWS, Azure and other transcription providers. Deepgram out does all of them by far.
Extremely accurate and fast responses
What do you like best about the product?
Deepgram determines context of a sentence as it's being transcribed and makes corrections as it receives more data.
What do you dislike about the product?
There's nothing I dislike so far. I am looking to complete a deeper integration across all of our services.
What problems is the product solving and how is that benefiting you?
Deepgram allows us to transcribe live streams and create captioning in real time.
It reproduces very fast and accurate result. And also it supports the real-time transcription.
What do you like best about the product?
High accurate transcript for general domain area even if it's doing with basic model without training. Speed is also fast enough to use in commercial product.
What do you dislike about the product?
Training is not easy to do. I wish it will support the self training in background, especially for language model. Hope to extend the API as well.
What problems is the product solving and how is that benefiting you?
Fast and accurate transcription for conversation in the contact center. It should support 8khz narrow band without media conversion. Deepgram is satisfied with this requirement.
Deepgram Reivew
What do you like best about the product?
I like how simple it is to use, how much flexibility it offers with speech recognition, and most importantly, how accurate the whole system is. I have to say it's the best speech-to-text I have ever used.
What do you dislike about the product?
Some improvements could be made as per the first initial steps, something to guide a user through the initial setup that's more visual and clear with step by step guidance.
What problems is the product solving and how is that benefiting you?
I'm trying to write a system to allow users to create websites purely using their voice. It's supposed to focus on accessibility for some who may struggle with typing keys.
showing 121 - 130