saras (Sanskrit for speech or anything fluid) wants to make pronunciation accessible and simple for those who are familiar with an Indic script like Devanagari (used for Hindi).

Learning another language is hard as it is and the fear of mispronouncing words makes new learners nervous.

saras not only bridges this gap, it can help someone speak almost any language (reasonably well) without learning the language; that too with no training! 😄

Give it a try, here.

Pronunciation of words is codified using the IPA extensively and can be found in dictionaries, but it is difficult to read without training. It is primarily based on the Latin script but it is a script unto itself, making it inaccessible.

Indic scripts however, are phonetic and have a rich coverage of sounds which make them very versatile and well suited for transliteration. They read like sheet music by design and are in daily use.

We stand on the shoulders of giants

• Wiktionary.com - APIs for IPA and the countless selfless contributors who created this, albeit imperfect but a rich resource for everyone.
• The sounds of English and the International Phonetic Alphabet
• ISCII - Indian Script Code for Information Interchange - Introduced by Bureau of Indian Standards (BIS) in 1991 as a superset for encoding Indic scripts. This was the precursor to its current home as a UNICODE block.
• LLM chatbots like ChatGPT, Github for hosting pages, search engines like DuckDuckGo, and by extension, the contributors on stackoverflow and the internet at large. They helped me learn, experiment, and iterate over with speed.