India’s population is fast approaching 1.4 billion. It has 22 official languages with more than 10,000 ‘unofficial’ languages.
With the launch of Jio, the largest 4G network in India, it is estimated that more than 829 million Indians will have affordable access to the Internet in two years’ time via smart phones that can be bought for as little as €39 ($43).
The expected growth in vernacular users of the Internet is set to grow to 90% of the population. Unfortunately, only 1% of digital content is currently available in any of those 22 official languages.
“English still dominates,” says Nulkar, “it is a languagechasm that needs to be bridged by facilitating the localization of English content, otherwise translation companies are going to miss out on a potential market worth an eye-watering estimated $7.75 billion.”
With all these figures whirling around in his head, Sandeep, who founded EUATC network member BITS International based in Pune, saw the huge potential. However, he also saw the problems. Firstly, the existing pool of freelancers and LSPs would be unable to translate the billions of words annually across all languages and this has led to many companies shelving any plans they may have had to localize digital content into vernacular languages.
“Another perceived problem to be addressed is that language quality and context cannot be balanced using current translation methods. At best it is patchy for modern approaches such as User Generated and Informative Content. Also, with a limited supply of traditional linguists and language companies, the ability to scale up and make localization affordable, looked to be severely hampered.
“A technical solution had to be the only answer. That is why I got together with an experienced team to co-found Vernac, a technology-driven platform that taps into the latent supply of educated bi-lingual Indian citizens empowering them to post edit content, from anywhere and whenever they are free and get rewarded for their work. In fact, we have deliberately gone down the ‘gamification’ route to harness the potential of millions of ordinary citizens,” he says.
In the first month since launching in a controlled environment and, only seeking Hindi/ English speakers, over 300 people signed up to use the Uber-style app.
“Vernac uses a mix of consumer and language expert reviewers, assisted by Translation Memories which, in turn, are linked to machine learning algorithms to deliver quality content in the right context.
“By tapping into millions of bilingual citizens, the scaling up drives the per-word rate down to less than one third of the current market price by stacking efficiencies and minimizing overheads like manual project management, claims Nulkar.
Supportive feedback is given to drive improvements and help the post editors understand where they stand in the ratings. Training is offered to help further drive the part-time bilingual linguists higher up in the ratings table, as this leads to better rates of pay.
Vernac has had a controlled launch. In its first year of operation it is aiming to process 65 million words, while in year two this will rapidly rise to 140 million. Based on current trends in the market Nulkar revenue will around $1.1 million by the second year of operations.
“There is huge potential with Vernac and we are seeking seed funding of $150,000 over the first year of operation with a view to acquiring resources, further development of the platform, building IT infrastructure and scaling up the operation as more of India’s official 22 languages are added to capability of the platform.”