The pivots that changed Uniphore

There are two pivots in the journey of Uniphore Software Systems, which, in the words of Umesh Sachdev, co-founder & CEO, have put the Chennai-based start-up firmly on the growth path. The first happened in 2014 when Uniphore, incubated at IIT-M incubation cell, got its first set of investors. That is when Uniphore transformed to a products company, launching three products targeted at different segments.

“That productisation worked out well. It was sane advice and we were able to show great scale in one year,” says Umesh.

It was thanks to this success, says Umesh, the company was able to raise its Series A round in 2015, when the next major transformation happened. The fresh set of investors, which included Kris Gopalakrishnan, co-founder of Infosys, and IDG Ventures, asked Uniphore to go global and also change from a licence fee-based income model to a SaaS-based (Software as a Service) subscription fee model.

“We went international. We went to Asia, West Asia, started going to North America, converted our entire offering to the cloud, SaaS,” says Umesh, 32. Switching to a SaaS-based model made it easier for Uniphore’s customers as they didn’t have to spend huge money on capex early on. “Moving from the licence model to the subscription was a good thing,” adds Umesh. It made the Series B round in 2017 easier, with John Chambers, former Chairman of Cisco, investing in the venture. Existing investor IDG Ventures put in more money and IIFL came in as an investor.

The pivotal point

For Umesh and his friend and co-founder, Ravi Saraogi, Uniphore was the second venture. Friends from Jaypee Institute of Technology, where Umesh obtained an engineering degree in information technology, they started a venture, Singularis Technologies, fresh out of college. That was a deep tech company, but the founders struggled to commercialise it. That was when they were referred to Ashok Jhunjhunwala, of IIT-Madras, in 2007. “That,” says Umesh, “was a pivotal point because Prof Ashok not only critiqued our business, but also gave us a path we could take.” Umesh and his co-founder decided to relocate to Chennai, from Delhi, and establish their new venture at the IIT-M’s incubation cell.

According to Umesh, he and Ravi told themselves that whatever idea they pick, it should impact billions of people.

“If this is going to be the idea that succeeds, let it have massive impact.” Their strength was in technology, telecom, mobile and signal processing, which is how speech recognition as a business happened. The incubation cell’s strong rural focus combined with their own strengths resulted in Uniphore, targeted at the rural population. They set up a call centre inside IIT-M that would provide all kinds of voice-based internet services to rural people. They went to places like Mayiladuthurai, Tamil Nadu and put up posters offering their services. In three months, they got about 10,000 calls, mainly with questions on livelihood, agriculture and government-related schemes. “We were convinced there was an opportunity. But we said we can’t scale this call centre model,” says Umesh. That is when they realised that there was this opportunity, but maybe the solution they were offering was not correct.

Therefore, instead of handling these questions through people at the call centre, why not have get computers to talk to the callers, giving rise to Uniphore’s current avatar of a speech recognition technology company. It was quite challenging then, in 2007-08, because there was no other company doing Indian languages and dialects. “We partnered with some who had cracked the code in English and European languages,” says Umesh, adding they decided not to reinvent the wheel, but to make it work for Indian languages and dialects and create IP around it. They approached enterprises and told them that Uniphore had this technology that could help them connect with their customers better.

In deep technology

“The company is still in the deep tech area of speech recognition, voice biometrics and text-to-speech. We have created multiple patents and IP,” says Umesh. He points out that, “as someone who started to solve the problem for rural India, we recognised that the technology has global application.” They initially focussed on 17 Indian languages and now do 90 global languages. From the beginning, they decided they would be a B2B venture. “We have remained true to technology, true to the direction, which is B2B. We have scaled globally. India is about 35-40 per cent, Asia another 40-odd per cent, North America at 15-20 per cent,” says Umesh. Over time, he expects America to contribute 50 per cent of the business.

Target segments

In 2014, when Uniphore switched to a products company, it decided to target three specific segments. The first was a speech analytics product, auMina, the company’s flagship product. This can be used in areas where it is hard to do something without technology, basically call centres, which handle thousands of calls.

According to Umesh, auMina listens to all the calls in real-time, converts speech to text, detects human emotions, detects voice modulation and because it is picking up all the calls, can do Big Data analytics, giving out trends to the call centre operators.

The second product was to automate something that was already happening. Not all calls to a call centre need to go to a human agent. So, this product, akeira, a virtual assistant, will answer queries from callers. The third area was to reduce possibilities of human error. “We said we should be able to use AI to reduce the chances of human error. That is where we introduced voice biometrics (amVoice). Because this is voice biometrics, this is a mix of hundreds of parameters, behavioural, physiological, it even distinguishes between a mimic,” says Umesh.

He says the opportunity for Uniphore to grow is enormous. There are over 6,000 global languages and Uniphore does only 90. It is expanding aggressively in North America. Umesh says Uniphore is the number one in its field in India and in the top three in Asia. “We clearly want to be among the top three companies in what we are doing globally, not just in India.” Speech analytics, which is AI solving an unsolved problem, is a $5-6 billion opportunity globally. Hardly any of this has been tapped. The virtual assistant, which is the automation opportunity, is much larger, at $30 billion. This too has hardly been tapped.

Switching to SaaS model

According to Umesh, the changes in business model in 2014 and 2015 were hard for the company, but Uniphore was able to handle the transition because the founders and the leadership team were hands-on and completely wired to the transition. When they switched to a subscription fee model, the company had to re-think revenue recognition. Similarly, converting to products wasn’t as easy as anticipated. The teams had to be split into product engineering and professional services. Now, firmly on the growth path, Umesh is confident of the future. He declines to disclose financials nor will he reveal how much the company has raised from investors so far.

It will use the Series B round to strengthen R&D and for sales and marketing in the US. It has a 110-member team now, which will over the next few years increase to 1,000.

Umesh believes that the company has enough funds to execute plans for the next 18-24 months.

“We are not in an industry where there are Goliaths and I am the David. It is anybody’s game. There are bigger companies, there are start-ups. But because it is deep tech, because it requires deep R&D, it is also not an area where hundreds of people would do it. A handful of companies in the world will play in this area. That is what gets us going. With great investors like John Chambers, IDG and Kris, with a great team that we have put in place and with the IIT Madras pedigree and now global presence, it is for us to lose,” says Umesh.

COMMENT NOW