This Saas-Based Startup Is Disrupting Call Centre Market With AI-Based Voice Bots
Entrepreneur's New Year’s Guide
You're reading Entrepreneur India, an international franchise of Entrepreneur Media.
Gone are the days when one used to spend 15-20 minutes to reach customer care executives over a call. Tapping the asked digit in a nippy’s time, later to get the call transferred to some other department, has been ruining the pre- and post-buying experience of customers for over a decade. The need for voice automation was inevitable.
The ongoing COVID-19 pandemic has propelled various sectors to install voice automation much sooner than what was on their road map to meet ever surging calls.
Vernacular.ai is one such company that is transforming the customer experience and engagement through its voice artificial intelligence (AI) solution. It has lately seen a huge spike in demand for voice automation solutions in both North America and southeast Asian countries.
Founded in 2016, this Bengaluru-based firm in May 2020 raised $5.1 million in Series A funding led by Exfinity Ventures and IAN Fund.
In an interaction with Entrepreneur India, Sourabh Gupta, co-founder and chief executive officer, Vernacular.ai shares the startup’s journey and how it is helping companies to provide better customer satisfaction.
‘Farmer’s Misery Was an Eye Opener’
Freshly graduated from IIT-Roorkee, Gupta and Akshay Deshraj, co-founder of Vernacular.ai, in 2016 came to Bengaluru with a dream to build a startup. They attended talk shows, events, met college seniors to get a grasp of what was in demand.
“During those times there was a lot of fuss about the next 50 billion people coming online and how there will be a surge in demand for financial products,” added Gupta.
However, Gupta and Deshraj had something else on their minds. They saw an opportunity in the upcoming 50 billion people because they knew these people will be different from those already online. To get a stock of the situation at a grass-root level, both went to Kanakpura, a city 55 km away from Bengaluru.
“While talking to a farmer one day, we got to know he used to travel 10-20 km to a bank’s branch just to understand a text message sent by the bank. In the process, he used to lose his daily wage for that particular day. This was a big shocker for us!,” Gupta exclaimed.
He said the next most significant challenge for the 50 billion people coming online would be a language barrier as all the interface exists in English, which is difficult for a country with 22-23 widely spoken languages.
They reached out to non-banking financial companies (NBFCs) and banks in tier I and tier II cities and understood they faced similar multilingual barriers.
This led them to formulate an AI-based voice assistant to help users. The startup initially started with chatbots, but soon realized its impact was not on a large scale and shifted to a voice assistant-based platform.
‘VIVA Enhances Customers After-Sales Experience’Vernacular intelligent voice assistant (VIVA)—Vernacular.ai’s proprietary—is an intelligent, multilingual platform that has been deployed in major enterprise contact centres to help boost customer stickiness and loyalty through a deep understanding of the customer's context and intent.
According to Gupta, businesses are growing, and so are their call volumes. However, they are always looking at ways to cut costs and call centres have always been expensive. From high training costs to paying salaries to thousands of employees, call centres have concerned companies for years.
“In order to cut cost, companies are pushing calls to interactive voice recognition (IVR), thus leaving no path for a user to reach an agent,” explained Gupta.
Enterprises are now placing Gupta’s VIVA in place of call centres which is cost-efficient and brings down considerable wait time, enhancing customer experience.
VIVA truncates the average call time of 10-15 minutes to 20-30 seconds. The AI- based platform can respond to both normal and complex queries.
Gupta said due to machine learning and vast database, the voice bot understands a user’s questions to reply back. The bot can also reach out to an agent in an extremely difficult situation.
The platform helps brands to boost their enterprise CSAT scores and brand loyalty.
To provide a better reach and bridge with the multilingual barrier, VIVA rides on natural language understanding (NLU) and speech recognition technology to understand different accents of users. Currently, the platform apart from English recognizes 10 different languages.
To gain expertise in various languages and build an understanding of different accents, Vernacular ai has been fed with data from various languages across the country.
“For a voice assistant platform, one has to solve problems in accents and dialects in geography like India, or else the product will not work,” added Gupta.
The platform with its cutting-edge technology through the voice can understand the age and sex of the user and accordingly answer. “If an older person is on the line then the platform will slowly read out the answer,” Gupta added.
‘COVID-19 Has Made Us Aggressive’According to Gupta, the ongoing pandemic has been a good tailwind for his startup. He said cost-cutting and social distancing norms together has made companies to fall back on AI-based chatbots to deliver customer satisfaction.
Before COVID-19, the firm generated 80 per cent of its revenue through the banking, financial services, and insurance sector (BFSI) sector. However, during the lockdown, the sector took a pause.
“We shifted our focus to the sectors that were growing exponentially during the lockdown. We focused on ed-tech, healthcare, DTH services because we understood the surge in call volumes they were facing,” Gupta added.
The firm which was growing 3-4 times in terms of revenue earning is now witnessing 5-7 times growth. It has also seen a 10 times growth in daily customer numbers. Currently, the company is working with more than 25 enterprises.The firm is also expanding in the US, Malaysia, Indonesia, and the Middle East. “We are going multilingual in the US as we have capabilities to understand American English and Spanish. Each of these languages has its dialects and accents. Our speech engine can understand these nuances and respond to people in their idiolect,” Gupta added.