Dr. Sumit D. Chowdhury, Founder & CEO- SesoVera and Chairman – Gaia explains the functioning of Hearz and how voice AI is emerging as a successful tech model
SesoVera Tech offers Hearz that provides a native voice platform for enabling offline & private multi-lingual voice engagement capabilities to things. It leverages Edge AI techniques using ultra low footprint ML algorithms along with advances in audio capture and rendering, and embedded coding to gather and process speech on mass-market inexpensive microcontrollers.
Understanding the Hearz Dynamics
Dr. Sumit K. Chowdhury, Founder & CEO- SesoVera has an interesting story to share on the genesis of Hearz. “We have been working with IoT and integrating IoT products into a variety of solutions. As we were working through the IoT solution set, we realized that there was a need and therefore huge potential for understanding spoken words and voices at the edge. We are not able to analyze it currently without using the standard publicly available technologies like Alexa or Siri or any of the public platforms. So we created a whole platform called Hearz, which we launched globally,” he reveals.
Hearz essentially ports the hearing ability in any object, be it a fan or a suitcase or a car door,a fridge and oven or anything, and can suddenly be able to not just listen but take action too. Hearz basically puts action and words together in IoT type devices. SesoVera’s journey started working with IoT, AI,ML on the edge and now they have put it together to create a very powerful platform.
Hearz is primarily a b2b platform player, with the customers mainly OEMs who are into manufacturing different kinds of devices. Hearz provides the full firmware, hardware as well as a platform for a private voice recognition platform, and this whole package is now ready for deployment. There can be numerous use case scenarios from white goods to vehicles and the automobile sector, to construction and manufacturing, to finally building and automation sector as well.
Informs Dr Chowdhury, “Essentially, we have seen a lot of people talking to us where they would require a voice based commerce model. And finally, we are seeing a lot of traction from healthcare and elderly care as well where we are able to provide alert based mechanism for devices that can be worn by all people. That is one of the use cases and the product I am ready to demonstrate.”
The Whys and Hows of Hearz
SesoVera’s go-to-market strategy for Hearz is fairly simple. It primarily focuses on its own sales channel and industry specialists who work in the OEM electronics space. It looks for opportunities where voice is going to be important, make a difference in the world and transform the way people interact with machines. “We approach those companies, show them our product and we request for a demonstration and an implementation opportunity with them. Once we show it and because our platform is very flexible, we are able to demonstrate to our customers very quickly the potential that we bring to the table and we subsequently work with them to integrate the product,” reveals Chowdhury. That has worked for us so far, according to the SesoVera CEO & Founder.
He agrees that a unique differentiator for Hearz over its competitors is the fact that it is a low cost producer of a very tiny microprocessor which can disrupt the entire spoken voice recognition industry. The other differentiator is that the platform is ready, while a lot of people are still working on it. The SesoVera team has been doing this for two and a half to three years now, and so its algorithms are quite trained. It enables them to deploy Hearz in a very short period of time into the customer’s devices. The speed to market of training an AI algorithm is something that is another disruptor for Hearz and that would differentiate it going forward as well to stay ahead of the curve.
The Importance of AI
AI is being used in quite a different set of paradigms right now. SesoVera’s parent company Gaia has been working on fusing AI to manage workforce, to capture the logic that their supervisors and managers use to route work, to allocate work to people, to manage and coordinate service levels that are supposed to be given. Corroborates Chowdhury, “In Sesovera, we are using AI to understand human voice and human speech and understand the context in which a person is saying something and then using that context to do some action. So we are a voice to action, using of AI to convert our voice into some action out in the universe.”
There are people who are using AI for navigation, people who are using AI for understanding perception of the world, people who are using AI for recognition of written patterns. There are a lot of applications of AI right now. The basic and the most commercial ones are the ones where we are eliminating repetitive work, which do not require much thinking on part of the doer, that is what is lending themselves into using AI right now where we can do it repeatedly and most accurately.
If we want to go to more complex forms of AI, there is always the issue of error. There is always the issue of a judgment, which may or may not be right. But, Dr Chowdhury advises, if you take the low hanging fruit where you take a piece of task, which is done repeatedly and you will find the logic with which the task is done and you programmatically manage that logic that is lending itself to AI. Lot of the RPA companies which are looking at the use of AI, are using it for that purpose. This basically implies there are a lot of disruption going on. It is not reducing the number of people required. It is actually making people more efficient.