The stars appear to be well aligned for Uniphore Technologies Inc. In mid-February, the Chennai- and California-based conversational automation platform, which provides companies with Artificial Intelligence-based tools for their customer service needs, raised $400 million (around ₹3,000 crore) in a Series E funding round led by New Enterprise Associates at a $2.5 billion valuation, catapulting it into the coveted unicorn club of startups with a valuation of $1 billion or more.
Uniphore, which last raised $140 million in March 2021, has raised over $610 million cumulatively till date. If these news items, however, give you the impression that venture capitalists (VCs) have always doted on it, or that the path has always been smooth for Umesh Sachdev, its CEO and co-founder, you would be mistaken.
Sachdev, 36, just wanted to be a successful professional. He was born and raised in Delhi and studied at the Delhi Public School, R K Puram. “My father spent the first part of his career with the public sector in the energy industry and eventually became a CEO (of a subsidiary of Tata Power called Power Lights) and joined boards of other companies. My sister followed in my father’s footsteps, did her engineering, did her MBA, and began to do well in her corporate career,” Sachdev explains. He adopted a similar path, opting for computer science at the Jaypee Institute of Information Technology in Noida, Uttar Pradesh, in 2003. “For me, undergrad was simply a stepping stone to success—do postgrad, then an MBA, and then corporate jobs. It was the only path I knew,” he says.
Life, though, had other plans for Sachdev. During the first year of his engineering studies, he met Ravi Saraogi who is now co-founder and president (APAC) of Uniphore. “Ravi was my batchmate. Being from a Marwari family, Ravi had only two options: either join his father’s construction business or do something of his own. Ravi had four years to figure out something.” After completing the first semester, Saraogi convinced Sachdev and a senior engineering student from the same college to collaborate on a summer project. “We were not going to get any grades for it. But Ravi said, ‘what are you guys going to do in summer? Let’s not waste time.’” Their ideation led to the prototype of a mobile location-based product, with the project extending to a full year.
As serendipity would have it, Paris-headquartered Gemalto—then the world’s largest SIM card manufacturer—held a technology competition around that time (Gemalto is now a Thales Group company). Sachdev entered the competition, and “we were in the top five teams of the world that they picked”. The trio made the trip to Paris, which was a “great recognition, but also the chance to visit Paris”, recalls Sachdev. Gemalto gave them a “very expensive software development kit (SDK)” and told them to complete their product. Even at this point, however, neither Sachdev nor Saraogi had decided to be entrepreneurs.
Work on the mobile project continued into their final year (2007-08). Around this time, Intel and the Union government’s department of science and technology (DST) were holding a business plan competition in India. Sachdev and Saraogi decided to develop a business plan for the product. “We applied and were selected in India’s top 10 teams. We were called to the Bangalore award ceremony.”
The next stage of the competition, a year later, was to be at the University of California, Berkeley, where the top three teams from India, Brazil and the US would compete. The Indian teams, which included Sachdev and Saraogi, were assigned mentors to help them refine their business plans, among other things. It was now that the entrepreneurial bug began biting them in earnest.
Around that time, their senior engineering colleague “decided that he would opt for a job”. Sachdev managed to resist the temptation of offers from companies such as Infosys and Accenture. His father too supported the idea, giving him more confidence. But getting money for the project was still a major concern. “In 2007-08, there was no venture capital in the country. We were trying to sell this product to telecom service providers and SIM card manufacturers,” recalls Sachdev. He even tapped into his father’s and college’s professional networks.
But there was one problem—the product was not selling. “We wondered what was wrong with our product, and why it was not selling.” It was then that someone suggested he meet a professor, the late M.N. Faruqui, at the Indian Institute of Technology, Kharagpur. He directed Sachdev and Saraogi to Prof. Ashok Jhunjhunwala at IIT, Madras. “That was a pivotal moment,” reminisces Sachdev.
Sachdev vividly recalls that “it did not take him (Prof. Jhunjhunwala) more than 15 minutes to tell us what was wrong with our business plan”. The verdict was clear: The current one wouldn’t work. Prof. Jhunjhunwala suggested Sachdev meet other startup CEOs at the incubator at IIT, Madras—one of them being a rural BPO (business process outsourcing) firm called DesiCrew. Inspired by the meetings, Sachdev and Saraogi were ready to pivot, and “started thinking of different ideas”.
Finally, the idea for Uniphore was taking shape. The idea was to use voice to enable access to the internet for the base of the pyramid population in India. “That was to be our company’s mission,” Sachdev recalls. Though Uniphore’s products today are powered by Artificial Intelligence (AI), Sachdev points out that “even when we raised the first round of financing in 2009, AI was not part of our pitch”. This early-stage seed round came from the National Research Development Corporation (NRDC), whose aim is to promote technologies and patents from universities and R&D institutions. “NRDC decided to invest back in the day what was equivalent of $100,000 (around ₹40 lakh back then).” It also gave Uniphore the liberty to choose their own board member. “So, all of a sudden, we had $100,000, we had a board member, we were a real company, and we were off the ground,” recalls Sachdev with a clear sense of pride.
Uniphore began selling the product to banks, telecom and microfinance companies with customers in semi-urban and rural areas. The proposition was to use Uniphore’s “voice recognition technology to build voice, mobile apps, voice banking, voice education, voice healthcare, and give it to people in different states because these are not the people who will come to your website anyway”. Uniphore, thus, became a business-to-business (B2B) product company.
But there was another pivot in the making. In 2014, an American bank got in touch with the startup. Sachdev wondered why, since they “don’t have any rural customer base in India”. The bank, it turned out, was interested in Uniphore’s voice recognition technology for its call centre. But this would also mean Uniphore would be “drifting from its mission of servicing the bottom-of-the-pyramid population in India”. Sachdev and team decided it would be worth it since it would expose their technology to a new set of users with an American English accent. “It (2014) was the first time we got a $200,000 contract.” Over the next two years, Uniphore began getting clients from BPO companies as well as banks in India, Singapore, even Dubai, keen to use their voice recognition product in contact centres. “We began developing voice, biometrics, voice authentication, and virtual assistant products.
“By 2016, Uniphore was a full-fledged voice AI contact centre focused startup,” recalls Sachdev. The idea, he adds, was always “to combine voice, natural language processing (NLP) and AI to basically make human to machine communication better”. During the pandemic, Uniphore extended the human-machine interface from voice to video and added computer vision to their AI model. “We have expanded our AI capability to not only understand words and sentences but also pay attention to what’s happening on the video in terms of facial expression and emotion,” explains Sachdev.
The strategy is paying dividends. Three years ago, Uniphore publicly stated that it was three years away from $100 million in revenue. “As we close to our fiscal on April 30, we would have achieved that goal,” insists Sachdev. At present, 70% of the revenue comes from North America; about 30% is from the Asia-Pacific region, which includes the Middle East, India, South-East Asia, Japan, Australia and New Zealand. This year, Uniphore expanded into Europe, “but it’s early days”, says Sachdev, adding, “and we are just about expanding into Latin America”.
Like most startups, however, Uniphore is yet to make a profit. “Our gross margin is trending up towards 80%, which is best in class to any SaaS (software-as-a-service) company. Our sales efficiency metrics are very healthy. So, in general, while we are not profitable yet, like any SaaS business, it’s very efficient. We are investing in new territories. We are creating new products. As that cycle of investment concludes in a few years, this (Uniphore) will be a cash-generating machine,” insists Sachdev.
To be sure, it has competitors: according to market research firm CB Insights, companies like Cresta, Cogito, Balto, ASAAP, Verint Systems and Five9, all of which use AI to help build AI-powered contact centres. That said, the global market has room for competition, with MarketsandMarkets forecasting that the speech and voice recognition market will grow from $8.3 billion in 2021 to $22 billion by 2026 on the back of an increase in the use of smart appliances.
Uniphore, on its part, has been acquiring companies to strengthen its AI portfolio. In October 2020, it acquired an exclusive third-party robotic process automation (RPA) licence from NTT Data. It bought contact centre automation software Jacada for an undisclosed amount last year. In January 2021, Uniphore acquired the Spanish video and emotion AI startup Emotion Research Lab. And, this month, it bought an AI-powered knowledge automation solution company, Colabo, for an undisclosed sum.
Going forward, Sachdev has two major goals. First, he wants to make Uniphore a $1 billion revenue company regardless of whether “it takes us four or five years”. The second is to go public “to provide liquidity to investors and its employees”. Sachdev, however, is hesitant to provide a time frame. He concludes, “I hope that by 2025 we are either very close or taking one of the two milestones that I spoke of.”