The Power of Speech
Two young entrepreneurs create a tool that breaks language barriers and closes the digital divide.
Umesh Sachdev was dejected and more than a little angry. This was 2007: He and his friend Ravi Saraogi had just presented an idea for their innovation, a tool that tracked mobile phones, to Dr Ashok Jhunjhunwala, a professor of electrical engineering at IIT Madras and head of the Rural Technology and Business Incubator (RTBI). Instead of an excellent review, what they got were ruthless remarks-their venture was not going anywhere, he had told them bluntly. What was most frustrating was he had broken it down, step-by-step. And his logic was irrefutable.
Today, Sachdev, 31, and Saraogi, 34, run Uniphore-meaning universe-a company built around speech recognition and its analysis, with over 70 enterprise customers and products that have touched the lives of over four million end users. Last year, Time magazine named Sachdev as one of the Next Generation Leaders who are changing the world.
Their beginning, as we know, was unremarkable. The two friends-graduate engineers from Delhi's Jaypee Institute of Information Technology-returned to base after their ruinous meeting with Jhunjhunwala. "We were disheartened at first: We've been working on this idea for two years, you've met us for the first time and so briefly, and all you do is reject the idea outright!" Over the next few weeks, they stewed over that first meeting, admitting, with great reluctance, that Jhunjhunwala was right after all. Sachdev and Saraogi contacted him once again, this time to dig deeper and seek mentoring. They were not going to give up on the idea of entrepreneurship so easily.
It helped too that the two did not have to look beyond their homes for role models-Saraogi came from a family of entrepreneurs. Sachdev's father and sister were corporate professionals. "Our families were encouraging and that helped us follow our dream. We wanted to do something unique and build something big," recalls Sachdev.
They went back and forth over the next few months, bounci...
Umesh Sachdev was dejected and more than a little angry. This was 2007: He and his friend Ravi Saraogi had just presented an idea for their innovation, a tool that tracked mobile phones, to Dr Ashok Jhunjhunwala, a professor of electrical engineering at IIT Madras and head of the Rural Technology and Business Incubator (RTBI). Instead of an excellent review, what they got were ruthless remarks-their venture was not going anywhere, he had told them bluntly. What was most frustrating was he had broken it down, step-by-step. And his logic was irrefutable.
Today, Sachdev, 31, and Saraogi, 34, run Uniphore-meaning universe-a company built around speech recognition and its analysis, with over 70 enterprise customers and products that have touched the lives of over four million end users. Last year, Time magazine named Sachdev as one of the Next Generation Leaders who are changing the world.
Their beginning, as we know, was unremarkable. The two friends-graduate engineers from Delhi's Jaypee Institute of Information Technology-returned to base after their ruinous meeting with Jhunjhunwala. "We were disheartened at first: We've been working on this idea for two years, you've met us for the first time and so briefly, and all you do is reject the idea outright!" Over the next few weeks, they stewed over that first meeting, admitting, with great reluctance, that Jhunjhunwala was right after all. Sachdev and Saraogi contacted him once again, this time to dig deeper and seek mentoring. They were not going to give up on the idea of entrepreneurship so easily.
It helped too that the two did not have to look beyond their homes for role models-Saraogi came from a family of entrepreneurs. Sachdev's father and sister were corporate professionals. "Our families were encouraging and that helped us follow our dream. We wanted to do something unique and build something big," recalls Sachdev.
They went back and forth over the next few months, bouncing ideas off Jhunjhunwala, forging ahead in spite of rejections-of which there were many. But they thrived on the candid, often brutal, feedback, taking it in the right spirit, realizing along the way that Jhunjhunwala was pushing them to think harder. "He told us how our ideas were all surface, we weren't going in depth … It was an important journey," Sachdev remembers.Finally, Jhunjhunwala suggested they move to Chennai to work under RTBI, a non-profit working with start-ups that service rural India, using information and communication technology. "It was our final shot," says Sachdev, "and we decided to give ourselves six months. Thankfully, we haven't had to look back since."
Ear to the Ground
They began by surveying the interiors of Tamil Nadu. Mobile phones had penetrated Indian households, but smartphones had not arrived then. They discovered that people in rural areas used mobile phones to make calls but not to access the internet or exchange SMS. They simply couldn't because feature phones did not allow vernacular scripts. And with over 780 languages and 66 written scripts in India, phones relied on English and Hindi. "We wanted to make a mass impact," says Sachdev. "So we started out by saying, 'let's get the 700 million people in our villages'-that's nearly 70 per cent of India's population." The market was untapped and the potential tremendous.
For their pilot project, Sachdev and Saraogi (then in their early 20s), put up posters across three districts in Tamil Nadu with a call centre number, where people reached out for information on agriculture, finance, health, education and entertainment. They received 10,000 calls in three months and realized that in order to succeed they needed to automate the system. "A computer would have to do the talking," says Saraogi, Uniphore's co-founder and now president, Asia-Pacific region. "After all, we were planning to cater to a vast population and the one-to-one call centre approach was not cost-effective."
Then, Uniphore was born in 2008. It implemented speech recognition based solutions to understand inputs from callers in their regional language and provide automated responses. "It is a tough science as people in our country speak so many languages and use multiple dialects," Saraogi explains.
Empowering the Marginalized
But it worked. So, for instance, Uniphore allows a Marathi-speaking farmer to speak to a virtual voice assistant and figure out the best market to sell his produce for maximum profits. The virtual assistant factors in his location, compares pricing and suggests suitable options. Various state governments are already using this technology to send multilingual voice messages to farmers with weather conditions and market prices. It also allows rural banks and micro-finance institutions to increase financial inclusion in the rural sector.
In the beginning, Sachdev and Saraogi contributed Rs5 lakh to the total seed capital of $100,000 (around Rs66.9* lakh at the current rate). The rest came from RTBI, Villgro-a social enterprise incubator that empowers rural development-and National Research Development Corporation.
Today, the company has three products. Akeira, a voice assistant (much like Siri on iPhone), supports 70 global languages (including 14 Indian) and 150 dialects. There is auMina, which generates most of Uniphore's revenue. It integrates speech and artificial intelligence to enable organizations to study and analyze all calls that come into their call centres, while offering customized solutions. Then there is amVoice, which offers voice biometrics solution or voice authentication.
"Now, Uniphore," writes Time magazine, is "helping hundreds of millions of people cross the divide between the digital and the real world by harnessing the power of speech". But Sachdev remains self-effacing. "Recognition feels great," he says, "but at the end of the day, we have to make sure that our customers are happy … we have to get back to doing what we do and get better at it." Meanwhile, Jhunjhunwala is utterly proud that his proteges have turned out to be so very successful.
*US$1 was Rs66.96 at the time of going to press.