Handcrafted content on wealth creation, entrepreneurship and more.

Backed by IDG Ventures, Kris Gopalakrishnan and other investors, Uniphore is building unique voice-based analytics and biometric solutions aimed at helping brands serve non-English speaking consumers better. 

umesh-2Think of a time when you’ve called a company’s customer service number to resolve a complaint or enquire about a service.  Often, before you are connected to a representative, a pre-recorded voice informs you that the call will be recorded for internal quality training purposes. But, how often do you think brands listen to each and every one of the thousands of calls recorded every day, draw insights from the conversations and make amends to their offering? The answer is, not often.

This is where one of Uniphore’s trademark solutions, speech analytics (auMina), comes into play. With the aid of technology, the analytics solution spots relevant keywords and detects emotions in voice conversations between the contact centre agents and consumers, to draw meaningful insights on how to improve overall call quality, enhance the agent’s sales skills, demeanour, knowledge and more. “Speech is a hard technical space. Although there are companies such as Google, Microsoft and Apple which are building solutions in the speech recognition and speech-to-text space (like Siri), there aren’t too many companies around the world, in the enterprise segment, which build solutions around this,” opines Umesh Sachdev, who co-founded Uniphore along with Ravi Saraogi.

While the company came into being in 2008, this was not the founders’ initial go-to business model. In 2006, passionate about creating an impact in the mobile technology and internet space, Sachdev and Saraogi built Singularis Technologies, a telecom-based company which can identify lost phones. But, after meeting several stakeholders and experts in the telecom industry, they realised that the model was not scalable. One among those experts was Ashok Jhunjhunwala, a professor at the department of electrical engineering at IIT Madras, who played a pivotal role in setting up the institute’s Rural Technology and Business Incubator (RTBI). “He was willing to mentor us under the condition that we relocate to his incubator inside the IIT Madras campus. There, he opened our minds to using mobile technology and internet to create an impact on the vast rural population in India,” recalls Sachdev. Thus, after twelve months of prototyping, partnering with Nuance, a global speech technologies firm, gaining mentorship from serial entrepreneurs such as PR Ganapathy (president of Villgro), Badri Seshadri (the founder of Cricinfo) and others, and raising a seed round (mix of equity and grant) of USD 100,000 from RTBI, National Research and Development Corporation and Villgro Innovation Foundation, Uniphore formally launched its first set of solutions in the market.


“As a company operating in a niche space, since there is no point of reference for the solutions you offer, when pitching to investors, you need to prove that the technology is mature to suit market needs, it is being adopted by some companies, and it has the potential to succeed in the long-term.”


Innovating on Product

To put it simply, Uniphore improves human-machine interaction by developing solutions to enable machines (computers) to understand and respond to natural human speech of non-English speaking consumers. In other words, it enables companies across sectors, such as banking, agriculture, FMCG and healthcare, to penetrate deeper into the rural markets, understand what motivates these consumers, and use those insights to deliver better services to them.

The company currently has three solutions on board; auMina (speech analytics), Akiera (virtual assistant) and AmVoice (voice biometrics). While the first solution, as mentioned above, derives meaningful insights from keywords and emotion detection to aid call centre agents, the second solution, Akiera, through its intuitive voice interface, handles unstructured consumer queries across several interaction channels. “It essentially delivers a self-service experience to the consumer, thus eliminating the need for IVR menus,” adds Sachdev. Lastly, AmVoice authenticates every user by recording their unique voice prints instead of requiring them to share sensitive data or passwords. This solution ensures that the authentication process is reduced to less than 15 seconds, and significantly reduces call time and operational costs. “We don’t have one single competitor who competes on all three solutions that we offer. Each of our solutions competes with a different set of companies, and even they are identifying different methods to solve the same problem,” notes Sachdev.

Until now, its solutions support over 30 Indian and international languages and in the founders’ words, have recorded a 95 per cent accuracy rate. But, when we say language-based solutions, we are exploring a vast set of dialects within the country itself. When asked how the founders’ identify which are the key languages they need to focus on, Sachdev shares an interesting identification strategy that they adopted. “We first partner with companies, identify which regions their consumers are based in and then develop language-based solutions targeted at them. This way, our solutions are focussed on specific client needs, and our impact is bigger,” he explains.  Some of the key clients they work with today include ITC Limited, IFMR, Axis Bank and American Express.

Additionally, serving in the niche B2B technology space can be a mean task for most entrepreneurs, and most often, the best selling point is the case studies they have developed based on the earlier clients they have serviced. Hence, when asked about the strategy they adopted to gain the first set of clients, Sachdev openly admits, “Prof. Jhunjhunwala played a key role in helping us build the right connections at the early stage. Once we onboarded the first set of clients through a known network, we used them as case studies to pitch to clients outside our circle.”

In fact, a similar strategy works even when pitching to investors, especially when a company is operating in a niche space. “Since there is no point of reference for the solutions you offer, when pitching to investors, you need to prove that the technology is mature to suit market needs, it is being adopted by some companies, and it has the potential to succeed in the long-term,” he says.

Apart from the grant the founders received in 2008, Uniphore had raised an angel round of US $1 million from IDG Ventures India, Indian Angel Network (IAN), Ray Stata, co-founder of Analog Devices and YourNest Angel Fund, and a Series A for an undisclosed amount from Infosys co-founder Kris Gopalakrishnan and existing investors (IDG, IAN, YourNest Angel Fund and Stata Venture Partners). The funds from the current round will primarily be channelised towards fuelling global expansions and improving product and core technology.

“We were successful in raising funds for our business because of several key lessons we kept in mind,” says Sachdev. According to him, firstly, an entrepreneur needs to put himself/herself in the investor’s shoes and understand what they are looking for. Secondly, you need to identify if your company is fundable. “Ask yourself this; is my company operating in a high growth sector? Is it serving a large market? Although there are many respectable businesses growing at just 30 to 40 per cent every year, they find it hard to raise money. It’s easier if you operate in a multi-billion-dollar space,” states Sachdev.

Looking Ahead

With an employee base of 90 across India, Philippines and UAE, Uniphore currently serves over 70 enterprise customers across key markets, and serves over four million end users. In fact, organisation culture being one of the key components to enable a company’s growth, Uniphore has introduced several interesting human resources strategies into its business. One among them is, to reach a 50 per cent female employee workforce in the next 12 to 18 months. “This is particularly a challenging task we have taken up, because the percentage of women as core engineers is lesser,” points out Sachdev. In other words, there aren’t many women in India and in global markets who opt for computer science as a field of study. Moreover, even at the hiring level, there is bias because people tend to subconsciously ignore the fact that there might be women as competent as men too, which means, unless hiring managers go out of the way to identify women with potential and skills in this field, it won’t be possible. “Hence, it is important for us to build a culture where we make our employees and hiring managers understand the value of creating diversity within the company. Only then will we be successful in fulfilling this goal,” shares Sachdev.

Currently, although there aren’t too many players operating in this space, Sachdev estimates the market opportunity to be US $6 billion to US $8 billion “And, since we have established that speech is the lowest common denominator for humans to interact with each other, as machines become smarter, speech will become a preferred way of communication with humans also,” he opines. As the company plans its foray into the US market, Sachdev sees its early move into this space and current traction in India and Asia Pacific as a clear lead to capitalise on the oncoming demand before other players do. “Our ambition is to try and create a fairly large global company in this space, like what Oracle built in the database and ERP space. We want to be the Oracle of our space with reference to the size and dominance they have now,” he says, on a parting note.

Madhumita Prabhakar
Madhumita is a Senior Writer at The Smart CEO Media Labs, the content creation partner for Beyond Basics@Wealth Advisors. Her focus is on chronicling the stories of India’s most exciting entrepreneurs from various sectors including technology, retail, financial servicers and education. For Beyond Basics, she’ll lead the effort to discover passionate social entrepreneurs who’re giving their lives to solve several societal problems in India.
Web Analytics