Features of CONVA

What enables our CONVA platform to add Retail Voice Assistants in under 30 min. The clock is ticking!

Pre-Built Voice Assistants

Domain specific User Journeys

We understand each domain has its own nuances. The Assistants support predefined User journeys specific to those domains.

Pre-Trained Speech Recognition

Higher accuracy through dynamic Improvements in Speech Recognition Models by training them according to the domain.

Pre-designed Conversations

Out of box Conversation Design designed for each user journey for each domain.

Pre-trained NLP Models

Our NLP models are pre-trained based on the SKUs and the nuances for each domain.

Multilingual out of the box

Speak in 5 Languages

We support 2 Indian languages out of the box - English and Hindi, with optional support of Tamil, Kannada and Malayalam.

Customise Translation

Have control over the words and phrases which you don’t want to translate. Bias our translation engine according to your app. Control what gets translated and how with customizable translation hints.

Automatic Translation

We reduce the overhead on your backend by giving the translated query from Indian Language to English.

Works with different Accents

Let your customer speak as they will, Slang is optimized to understand different accents.

Intuitive experiences

Helping hints

We provide rotating hints on the Slang Surface. This helps in guiding the user of what they can speak thereby reducing cognitive overload. But Dil Maange More? We even have a hints button to show even more hints

Patience Mode

We understand people take time to speak, so we wait for the person to speak before zonking out. Did we mention designed with empathy already?

Talkative Mode

Some people speak more than others(<emoji>Our CEO</emoji>), we won’t stop listening to the user if she is in the middle of a long sentence

Talkback Mode

A conversation is two way. The Assistant will ask questions and also talk back to your customer in the language of their choice.

Mute Mode

Sometimes you want to talk but not the Assistant. Fear not with the mute mode option and the Assistant will only listen but never talk back

Different Triggers

The Assistant can be invoked either via an inline trigger embedded inside the app or via a global (default) trigger that floats on top of all screens automatically

Slang Labs Logo

Slang Labs provides accurate and multilingual voice assistants. These in-app intelligent virtual assistants can be used out of the box for android and web apps with just a few lines of code.

Find us online