[Tech30] With Navana Tech, these brothers are building a mobile SDK to make the Internet Bharat-friendly

With their litOS (Literacy OS) built at Cornell Tech, Raoul Nanavati and Jai Nanavati found the roots for Navana Tech, which looks to provide the right UI for non-English speaking internet users in India

By the early 2000s, with many households having computers, a majority of Indians had learnt the basic operations of a computer. As technology partnered with convenience, mobile phones and free data reached our fingertips, and so did a plethora of information and knowledge.

It is convenient to imagine that digital operations are child’s play in today’s world, considering how acquainted most of us have become with the internet and digital devices over the last decade and a half. But, there exists a large part of India which is introduced to the digital ecosystem with little or no prior hands-on experience with smartphones, data, and the like.

It is to solve this problem for the next billion users of India, especially the illiterate sections of the population in semi-urban and rural areas, that Raoul Nanavati (31) along with his brother Jai Nanavati (28) started Navana Tech in Bengaluru in 2018.

It is an artificial intelligence (AI) platform that builds UI based on speech recognition technologies that actually speak and listen to users in the native languages of the country. The machine learns and trains its own model further by understanding the nuances of the users’ dialects, to serve them better.

The team at Navana Tech. Raoul Nanavati (extreme left), Jai Nanavati (extreme right).

Solving a pain point

Back when mobile data penetration had its inception in the country, nomenclature like ‘the first million/billion internet users’ also came on its own. Around that time, Raoul was building mobile games for Indian B2C users while running his last venture BYOF studios out of Mumbai.

"Most people would think of adopting best-case practices from the West, like building a Candy Crush for India. But, that doesn’t necessarily mean success for our market. Variables like language and text dependency come into play, since most Indian users cannot speak English," says Raoul.

Raoul also realised that users had to make a decision after reading through a lot of text in English. Also, if they had to navigate through a series of layers to finally arrive at the desired main screen of the app, the multiple-layered information architecture came with a lot of text.

The real problem arrived when they had to return to the main menu after reaching the end window of the interface flow. Many users were dropping off from using the app due to their inability to process the information provided in English, and the retention metrics at BYOF studios took a major hit as a result.

Raoul says that he was able to identify a lot of patterns in how UI is usually designed, which were all connected. He understood that these might be the potential user interface (UI) problems for the future billion users, and hence started Navana Tech.

How does it work?

The startup leverages text-to-speech and audio files for voice-based assistance and to create illustrative iconography. The flagship product, Bharat SDK, enables smartphone companies to implement a text-independent, visual, and voice-driven user experience on their apps.

“The preliminary thought was to create a design paradigm to actually create the most possible effective product on smartphones which would have the highest impact for first-time Indian users. We realised that we had to create an operating system and sell it to OEMs (original equipment manufacturers).”

The startup’s solution is essentially a layer-on-top, which requires zero changes to the existing app interface. It combines three modules ⁠— contextual voice assistant, voice-based chat interface, and illustrative iconography. The languages currently on offer are Hindi, Tamil, Kannada, Odia, Telugu, Bhojpuri, Bengali, Marathi, and Gujarati.

The startup’s B2B customers like Ujjivan Small Finance Bank now use Bharat SDK for low-literate users. The enterprises are supported in adding functionalities to the assistants that animate the buttons, which can explain these functionalities to users in their local languages.

The team

Raoul was previously the head of product strategy at WOI and the CEO and Co-founder at BYOF Studios. He has a BA in Political Science and Economics from McGill University, Canada. Jai was also an enterprise software engineer at WOI as well as at Elemential Labs. He has a bachelor’s in engineering from the University of Michigan.

Both the brothers went to Cornell Tech University in New York for their MBA, where they developed litOS using agile methodologies and worked in weekly sprints. They frequented in gathering user feedback, to constantly tweak and improve the user interface.

At the university’s startup awards competition, the team bagged $80,000 in funding to build the product for Navana Tech.

The model and the market

Navana Tech’s SaaS-based model makes revenue from lead generation. The startup charges the enterprises on a per-transaction basis and gets paid only when users get through the flow till the end successfully. To date, the nine-member team has onboarded 4 paying customers.

With tech giants like Amazon rolling out Alexa, voice recognition technology has gained a tremendous amount of attention for business use-cases in the Indian startup ecosystem. Also, with vernacular content becoming a focus area, startups like Vokal, Gnani.ai, Liv.ai, Reverie Tech, and Vernacular.ai are rapidly building their automated speech engines for the next billion users of India.

The potential of this sector is borne out by recent deals, with Reliance industries acquiring Reverie Technologies for Rs 190 crore and Walmart subsidiary Flipkart acquiring Liv.ai. According to a report on the speech-and-voice-recognition market, the voice technology market is expected to grow from $7.5 billion in 2018 to $21.5 billion by 2024, at a CAGR of 19.18 percent.

The expected growth can be attributed to the technology’s potential in healthcare applications, the growing demand for voice authentication in mobile banking applications, the rapid proliferation of multifunctional devices or smart speakers, and the growing impact of AI on the accuracy of speech and voice recognition.

Right now, the team is exerting its focus on perfecting its minimum viable product, so as to test the possible use-cases with how they can collaborate with B2B users.

(Edited by Athirupa Geetha Manichandar)

YourStory’s Tech30 companies list is an annual selection of 30 carefully curated and disruptive tech-based startups that we believe will shape the new narrative for India and the world. To get a complete overview of the 2019 Tech30 companies list, download the Tech30 Report here.


Updates from around the world