+ What is ajala.ai?

ajala.ai is a platform of speech technologies and tools for automating voice experiences in African languages. Our primary focus is enterprise solutions.

+ What languages do you support?

We focus on broad coverage and support for African languages. Please see our language coverage map for an up-to-date summary.

+ What languages are in your pipeline?

Our language coverage map summarizes current service and language coverage. We aim to provide broad coverage and support for African languages, and currently have about 20 languages in our development pipeline. To discuss your enterprise needs, please send an email to support@ajalaco.com.

+ Do you have any other speech or NLP services in your pipeline?

Our primary focus is in speech technologies for voice automation. We currently offer voice recognition and speech synthesis in various African languages. We are working on expanding our speech technology offering to include additional services, and will also offer NLP solutions that support voice automation solutions. For updates on new product releases, please sign up to our newsletter.

+ Do you have a demo I can use, or interact with?

A recording of our launch event where we demoed our services is available online. For a demo, please create a free account and book a meeting with a sales consultant.

+ What enterprise integrations do you offer for voice automation?

We currently provide integrations for Cisco and Avaya enterprise IVR solutions. We also offer solutions that support integration with cloud telephony platforms including Twilio and Nexmo. For more information, please get in touch.

+ Do you offer on-premise solutions?

We provide containerized on-premise and hybrid deployments for customers who cannot use our cloud platform because of specific data privacy/security restrictions. For more information, please get in touch.

+ I’m interested in voice recognition/speech synthesis for African languages. Can I test your service?

Thanks for your interest! We are currently in private beta and our platform is only open for testing on a limited basis. If you represent an enterprise looking to develop voice automation solutions, please email hello@ajalaco.com to upgrade your demo account to access our services. A recording of our launch event where we demo the platform is available online. To learn about new releases and platform updates, please sign up to our newsletter.

+ How accurate are your voice recognition models?

Accuracy depends on a number of factors including the frequency of the audio recording, presence and loudness of background noise, speaker’s articulation, speaker’s distance to the microphone/recording device, and presence of multiple speakers/multiple languages. In “ideal” circumstances, our voice recognition models achieve recognition accuracy rates ranging from 80-97%, depending on the language.

+ My company is interested in leveraging voice automation to deliver an enhanced customer experience in the African market. How can ajala help deliver our goals?

We would be pleased to help your company in this journey! We offer technical, operational, and market consultancy to help optimize voice as a channel for delivering customer value, driving revenue, and reducing cost. For more information, please get in touch.

+ Does my organization need to automate all of our voice customer service interactions to benefit from ajala’s platform?

We believe in a hybrid approach to developing bespoke voice automation solutions tailored to the African market. A mix of human and automated customer service solutions can improve the economic and operational efficiency of your organizations customer service pipelines, and open new revenue opportunities.

+ How do I register for an account?

You can register for a free account here.

+ I’ve lost my account password. How can I recover it?

Please raise a support ticket and include your username.

+ What does your service cost?

Platform costs are billed on a volume-of-use basis for each service and language your business needs. Our pricing provides businesses pay-as-you go flexibility, and control of service and language access. Higher volumes of use are awarded discounts. Demo accounts are provisioned free, limited access for testing purposes only. For more information, please get in touch.

+ How will I be billed for using your services?

Customers are billed on a monthly basis, based on service usage. For more information, please get in touch.

+ How do I upgrade from a demo account to a paid account?

To upgrade your demo account to provision access to specific services and/or languages, please get in touch.

+ I am a company interested in partnering with ajala. How do I initiate a partnership with ajala?

Thanks for your interest! For companies interested in discussing potential partnership opportunities, please get in touch.

+ I am a vendor interested in distributing ajala’s solutions. Do you have a vendor distribution program?

Thank you for your interest! For vendors interested in re-selling and distribution opportunities, please get in touch.

+ How is data posted to the service to execute requests stored/used?

Ajala’s speech technology services automatically log all requests and responses in an anonymized manner as part of overall system monitoring. Such data is not made public but may be used to implement improvements to individual services. Information is stored in a manner that supports the European Union’s General Data Protection Regulation (GDPR), and complies with ajala’s security and data privacy protocols.

+ How secure is the platform?

Ajala supports TLS encryption to protect communications between Ajala and application external to the platform. Use https: API URLs to employ encryption. Requests are authenticated using individual API keys which should be kept private, secure, and shared only with relevant parties.

+ How can I execute a voice recognition request?

Our documentation includes a tutorial to help you execute your first voice recognition request.

+ How can I execute a speech synthesis request?

Our documentation includes a tutorial to help you execute your first voice recognition request

+ I created an account and have an Auth Key, but when I make requests I get a “service unavailable” response.

When you register for a new account, you are provided a demo account that has limited service access. To request access to specific services and languages, please send a support request including your username.

+ Sometimes, when I issue voice recognition/speech synthesis requests, I get a “No workers available” message. What's going on?

While we are currently in private beta, and working to scale platform resources to meet increased demand. This message is likely an indication the platform was experiencing high demand at the time of the request. Please send a support request along with the request, time of request and username to help us assess platform usage requirements.

+ What are your usage requirements?

Ajala’s cloud platform can be accessed by any third-party client that has Internet access, supports encryption, and can interact with a RESTful API. Local data and storage requirements depend on the nature and design of the application making the requests. For on-premise deployments, please get in touch to discuss your use-case and technical requirements.

+ How can I improve the accuracy of a voice recognition request?

Accuracy depends on a number of factors including the sample frequency of the audio recording (the higher, the better), presence and loudness of background noise, speaker’s articulation, speaker’s distance to the recording device, and presence of multiple speakers and/or languages. Telephone calls are always recorded at 8kHz, however, third-party applications that do not rely on audio generated during a telephone call generally generate audio recorded at a higher sample frequency. Audio recordings that optimize the factors described above should achieve improved recognition accuracy.

+ I noticed one of your TTS voices mispronounced a word. How can I fix this and/or notify you of this?

The platform currently supports a common pronunciation model for each language in our offering. To report an incorrect pronunciation, or a pronunciation variation you would like us to support, please send a support request that includes the details. We are working on introducing support for bespoke pronunciation models. For updates on the platform, please sign up to our newsletter.