MAKD has very good experience in collection custom speech data using our clients' tools to help voice recognition systems. We greatly focus on the quality of the recordings as we use only native speakers located in different countries around the world. 

We at MAKD follow special steps to ensure that everything is going well. After agreeing with our clients about the work frame and studying the instructions and guidelines very well, we provide training sessions with very clear instructions to our collectors using different methods to make sure that everything is very clear. We do care about the gender balance, required age and to have unique people to participate in the project. We work very closely to our clients and our people to make sure that the recordings meet our clients' specifications. 

In addition to the collection, we also have very well trained teams to do the validation part if our clients need this to produce very accurate recordings that help in Speech recognition machine learning very easily and effectively.

We have worked on collecting thousands of hours of speech data collection in different languages and dialects such as: Arabic, with its dialects, English, Hindi, German, French, Turkish, Indonesian and others. We did the collection part in addition to the validation part for those data collection projects.