Due to COVID-19 Scenario, 80% businesses want to localize their content-  Mr. Arvind Pani, CEO & Co-founder, Reverie Language Technologies

India has 560 million internet users, and the number is growing because of easy access to smartphones and the internet. This makes it the second-largest online market after China. By next year, we estimate about 500 million Indian language internet users to be online apart from the 200 million Indians who access the internet in English. Hence, it is really necessary that the Indian internet also has an adequate amount of content in Indian languages, which today is only 0.1%.

In order to establish India as a self-dependent economy, Reverie Language Technologies has developed the solutions to close the language gap in the country: Prabandhak, a unique cloud-based, AI-powered translation management hub that ensures swift, easy, and accurate localization. It helps to organize work, automate, and expedite processes while managing your multilingual content – all on one efficient platform. Prabandhak also has features such as an in-built marketplace that enables businesses and individual translators/LSPs to discover and interact with each other in real-time depending on the set budget, quality, and time expectations.

And the second solution the company offers is Anuvadak: It is a platform that accelerates the process of creating, launching, and optimizing your website in multiple languages. The platform enables you to connect with customers in their language with faster go-to-market and effortless content management. Anuvadak can scale down the website localization time by 40% and can save as much as 60% of the localization and content management costs.

Mr. Arvind Pani, CEO & Co-founder, Reverie Language Technologies with whom we had an email interaction to know more about company insights and their upcoming plans, and we glad to share the same here with you all netizens.

 

What inspired you to lay down the foundation of Reverie Language Technologies?

The year 2009 saw the inflection point for the adoption of smartphones in India. People were beginning to be more receptive to smartphones and converting from feature phones.

We saw that the mobile phones didn’t even have the basic capabilities to even properly display Indian language characters. They would render as junk characters. We saw a huge gap here and founded Reverie Language Technologies in 2009 to solve Indian-language computing problems on mobile phones.

In 2011, when we partnered with Qualcomm after having won the 2011 QPrize, we were the first company in the world that offered Indian-language computing solutions for all 22 scheduled languages of India.

While the mass media like TV, radio, newspapers have figured out this difference and have localized their services in many Indian languages, the internet has been extremely slow to catch up. The world has moved online for the majority of business services and products. However, the internet has been serving only the minority of those who can understand English in India (less than 10% comprehend English in India), leaving out the majority.

Our foundation is set on a vision to make important information more accessible to all of India, in Indian languages. Simply put, bridging the language divide on the internet inspired us to found Reverie Language Technologies.

 

How is Reverie Language Technologies different from already existing language translator programs?

Our translation technologies are trained on industry-specific vocabularies and corpuses. This means that we are capable of producing highly accurate, business-quality translations for a particular vertical or domain.

We offer automated grammar checks and error checks in the output, such that the accuracy, quality, and text format conversions (from source to target languages) are preserved.

Our language solutions take care of uniformity across business phrases, terms, and terminologies, preserving brand consistency across various translations.

We go beyond translation, however.

Reverie Language Technologies is the only company that offers end-to-end language technology products and services in this space for Indian languages. We offer a wide range of language technology products for Indian languages for both text and voice solutions, including:

• Machine translation and transliteration technologies specifically trained on and built for Indian languages,
• Website localization,
• Scientifically designed multilingual text display suite,
• Industry-specific Indian-language vocabularies that help with accurate business translations,
• Easy input products like multilingual keypads that support phonetic typing,
• Multilingual search and accurate next-word predictions,
• Automated speech-to-text and text-to-speech conversion and more.

Our translation technologies power a whole range of digital assets, such as websites, applications (mobile/web), chatbots, communications (SMS, push notifications), IVRs, chatbots, voice bots, and more. In fact, we are the only Indian company that offers end-to-end products in the Indian-language industry market.

Another important factor to consider is that we have a strong technical understanding of the complexities and nuances of Indian languages.

 

The big problem while translating the content from one language to another is the accuracy, how Reverie is handling that, and what about its accuracy rate?

With the same data and similar algorithms, different companies produce different accuracies in the translation outputs. We provide the highest accuracy in the market by preprocessing data and using algorithms specifically for Indian languages. This way, we get a better output in comparison to others. We essentially exploit the properties and nuances of languages and our tools understand these.

Our translation technologies are trained on industry-specific Indian-language vocabularies and corpuses, across both text and speech. This means that if you are in the banking industry, we can tweak and train our AI-powered engines for text and speech to understand and learn banking terms, terminologies, and context, thereby delivering accurate results.

 

Could you please elaborate, how exactly Reverie works to translate the content of apps or websites?

Prabandhak, our AI-powered, end-to-end translation management hub is powered by Reverie’s Neural Machine Translation (NMT) engine. The platform offers specific workflows for automated translation, human intervention, and evaluation. The NMT engine is integrated with Indian-language specific vocabularies, translation memories, glossaries, dictionaries, and more to automatically translate English content to multiple Indian languages. Refer to question 3 on how these technologies are trained to improve accuracy and relevance.

Anuvadak, our website publishing, and managing platform enable businesses to accelerate the process of creating, launching and translating existing and new websites in any language, be it Indian or international, coupled with either Reverie’s NMT or one of their choice. The platform allows for manual verification and management of data as well. Any changes made in the source language are automatically tracked and reflected in all the target languages. Further, the target language translated websites are SEO compatible in multiple languages, crawlable and searchable by search engines, improving their online visibility and discoverability.

Further, our voice solutions (text-to-speech and speech-to-text) are trained on voice models of different languages and dialects, powered by industry-specific vocabularies, to simulate human voice output and to understand voice inputs of different accents.

 

Please enlighten us with Reverie plans or just give a rough idea, how much one needs to invest to get started with Reverie?

Our price is directly proportional to product usage. Therefore it is affordable.

 

What is your future road map for the next 2 years?

We are focused on building products to address all user engagement aspects, be it input, search, voice, translation, or localization. We are also focused on improving accuracies and relevance of our language models to improve the overall local-language user experience.

We plan to empower more number of the rapidly rising Indian-language users with our language by enabling large businesses and governments to connect with more people in regional languages.

 

Any plans to support other languages of the world apart from India’s?

Anuvadak is a language-agnostic and enables you to publish and manage websites in any language with a corresponding NMT of your choice.

Our other products are Indian-language specific. There are 22 scheduled Indian languages with their own unique complexities and nuances. This itself is a huge problem to solve, and there is a lot to do in this space. Therefore, we first aim to solve Indian language computing problems to serve our Indian-language communities.

 

As India is a multilingual country, therefore as per your view, localizing an app or website brings what kind of benefits, growth, or revenue to a business?

Localizing business and government websites and communications brings a lot of benefits at various levels:

Shift in the focus from revenue to inclusivity: Ideally, digitizing your physical store presence involves emulating diversity and inclusivity online as well. Currently, the internet serves only English-speaking users. Your physical stores serve all kinds of customers irrespective of their language. This barrier needs to be first removed. Localizing any digital presence should be focused on inclusivity, especially in the pandemic era where more linguistically diverse users are flocking to the internet for services and information.

Meeting rising rural demands: According to the latest May 2020 IAMAI report, for the first time in India, rural internet usage has surpassed that of urban. As of Nov 2019, according to the report, there were 227M rural internet users as opposed to 205M internet users in urban areas. That’s a 10% difference. This difference is only going to increase in the next few months to years. If you don’t localize, you will be left behind. Rising demand for local-language content is no longer a hypothesis. It is a fact.

Turning new users into paying consumers: This new Indian-language user base is actively looking for online services, especially in the wake of the pandemic. Be it banking, or eCommerce, or delivery services, local-language users are looking to consume these services. However, the bottleneck is the language problem. They can’t use services they can’t understand.

Boosting the bottom line: Engaging with this underserved audience would directly boost any business’s bottom line. According to a KPMG report, more than two-thirds of local-language users trust digital content in their languages to be more reliable than that in English. A vast, new user base and renewed trust mean more revenue.

Access to more information for more people: Less than 10% of Indians understand English. Localization of important information like government schemes, rights, product and service descriptions, scientific information, etc., into many Indian languages, would mean rising awareness and education among Indian citizens.

Raising India’s economic growth: Engaging and educating people in the language they understand will in turn create more innovators to be seen and heard. A lot of new jobs can be created in the language tech space alone. Machine translation still needs human intervention, and we could use talented local-language translators to improve machine translation.