"The development is being done as we speak," Rohan Jayaweera, country consultant for US-based Google Inc told an LBR-LBO Summit forum.
"I am sitting in a panel with a number of Googlers across the world where we are refining the algorithm."
Jayaweera a "tremendous amount of data sets" in the Sinhalese language in useable fonts was needed to develop the translation algorithm.
Sinhala is spoken by the majority community in Sri Lanka and is considered to be part of the so-called Indo-Iranian language family. It uses a variation of a Pallava-style alphabet, used in India and some parts of South East Asia.
Google Translate allows writing in one language to be translated into English or any other language. It can also detect the language.
"Now that we have collected a significant amount Sinhala data - Sinhala words actually - the next challenge is alignment," Jayaweera said.
Alignment involves working on the grammar and syntax of Sinhala.
"So we are very optimistic we can launch it very soon," Jayaweera said without specifying a data.
He said even after a launch "a significant time was taken for iteration and refinement."
"It wasn't as easy where we did have millions and millions of words that we need to be able to identify and connect," Jayaweera said.
Translation allows knowledge to pass among human being very quickly, breaking nationalist barriers.