Conversational AI will change the way people interact with devicesVishal Krishna
Voice and chat based conversations are entering the realm of AI where software will figure out patterns and solve needs.
The past week has been a great one for Haptik.
It won an award at the Amazon AI conclave and while speaking at Mobile Sparks 2017 Aakrit Vaish, CEO and founder, enthralled audiences on how contextual AI was becoming relevant regardless of the device that a person is using.
He said that the world of voice and chat were converging where AI would figure out the nature of the conversation.
This, he explained, was possible if divided into two formats. Need-based conversations and context would be solved by voice. For instance, if you need to order your vegetables or grocery list, this would be solved by voice.
Now say you want to understand how to fix the plumbing in your house, this conversation would be a chat interface with the software or a combination of both voice and chat.
“Organisations and humans are on a similar path. But one needs to figure out the context in which they will communicate and we must build IT,” explained Aakrit.
He explained there are five things that work for AI has a conversation interface:
- If messaging is where you spend time then platforms like WhatsApp or WeChat will power AI.
- From these platforms great content can be built.
- There is power customised notifications.
- Simple tasks don’t require apps or websites. Build conversations for this.
- When the above happens you can have chatbots show up anywhere.
“Existing distribution platforms like WhatsApp and WeChat are opening up and developers can build AI on top of them,” said Aakrit.
Today Alexa of Amazon is doing well and millions of enterprise devices are connected across the world. “It will be interesting to see how Amazon integrates Alexa on its e-commerce platform,” he added.
According to him, voice-based AI will be used in tasks that are repetitive. Chatbots or text-based will be used in banks, manufacturing, and several other business use cases.
Haptik has 27 customers in total and is involved in solving AI in Indian languages. “Today we are skydiving and have just balanced ourselves in the air. As Haptik we have a great view of the surface and we have to find those great business avenues that want to use AI,” says Aakrit.
So developers must keep in mind that AI is about context and not about device or form factor.