Manoj Annadurai wants to reach out to the huge untapped markets of rural India with his Indian-language software

Indian software house Chennai Kavigal has developed regional language office productivity software on the GNU/Linux platform. Frederick Noronha finds out what motivates the company and what this initiative could mean for the spread of computers in non-English-speaking India

Chennai Kavigal’s Shakti office suite offers common office productivity tools on Linux in regional Indian languages, at an affordable cost

It looks like MS Word, but it’s in Tamil. You could do your spreadsheets in this software, and it works in Hindi. Thanks to the initiative of a young and enthusiastic firm in South India, useful office-based computing solutions are finding their way to Indian-language computer users.

Going by the atypical name of ‘Chennai Kavigal’, the firm has been toiling to provide everyday office applications in Indian languages. “We started with word-processing and now have an entire office suite. We’re working on handwriting recognition, and porting these options to (the ‘free’ operating system of) GNU/Linux,” says CEO Manoj Annadurai.

Such a solution holds out exciting possibilities, and not just because the average computer user badly needs ‘office’ tools, that allow for commonly used computer tasks to be carried out. More importantly, such tools are sorely lacking in many Indian languages, at an affordable price.

Based in Chennai, the firm Chennai Kavigal’s product ‘Shakti’ seeks to provide the functionality of Microsoft Office. But, in Indian languages. “You don’t have to buy a suite of seven different applications. This offers all the products in one,” says Annadurai. It currently comes in Hindi-English and Tamil-English versions.

With the product priced at Rs 1,995, the company seeks to offer an affordable product to Indian computer users who would be more familiar using this tool in their regional language.

Chennai Kavigal has a staff of about 40. “We’ve got a good response for this product,” says Annadurai, a mechanical engineer. The company is already working on a version in Telugu and plans to launch Marathi, Gujarati and Bengali editions shortly. “It doesn’t come with all the bells and whistles of MS Word. But this is a functional word processor,” explains Annadurai, pointing to one of his products.

A regional alternative
What’s the logic of going about re-creating office productivity tools in Indian language versions?

Argues Annadurai: “Linux is actually expensive. In one sense, it is free (in the sense that it is actually freely copyable, without licences needed to be purchased for each user). But to get persons trained on Linux is tough. Today, even children are using Windows. If you want a police constable to be using a computer, convincing him about Linux is going to be tough.”

On the other hand, legal copies of Microsoft applications don’t come cheap, in an Indian context. While the Windows operating system itself isn’t overly expensive, a package like Microsoft Office could be priced at anything up to Rs 25,000.

“IIT-Madras and Chennai Kavigal wanted to do something about that. We wanted the entire office-tools to be available to the Indian language user, at an affordable cost. In a language that they found useful,” says Annadurai.

To make this solution useful, it should ideally be bilingual, if not trilingual. Ideally, it could include English (the ‘colonial’ language, but still widely used to interconnect different parts of the nation and known by most who have gone in for a higher education), Hindi (the ‘national language’ and widely spoken in North India) and the regional language.

Secondly, argues Annadurai, to make this useful and relevant to the common man, the software needed to be inexpensively priced.

Complicated process
Getting regional language solutions in India is not easy. Indian language computing has its own complications. Keyboards and computing has been geared to meet the needs of a 26-alphabet English language. Indian languages have a few hundred characters and character-combinations.

“Mapping this onto a keyboard meant primarily for 26 alphabets, means that one needs multiple keystrokes to get one character. In a language like Telugu, you could need four to five keystrokes to create one character,” says Annadurai, adding that his company is working on a pen-interface for handwriting-recognition systems.

Innovative company
The firm’s name Chennai Kavigal simply means Poets of Chennai. “We did not want a name with Microsystems or Macrosystems, but wanted something that was Indian. If Panasonic, Matsushita, Daewoo, Hyundai and Sony can be accepted, why not this?” asks the CEO defiantly.

Chennai Kavigal was started at the end of 1996.

“If recognition means that a company’s product is bought by a lot of people, then a lot more needs has to be done by our company,” the CEO admits. “But unlike other Indian companies that are more into services, we believe in creating products. That is what would make India a software superpower. Sun, Oracle and other big names have their own products. But which Indian company has put out recognisable products, apart from a few accounting packages?”

The company’s future plans include alliances with bigger companies, which could give its products the badly-needed wider exposure. Also, Chennai Kavigal is eager to build deals with the government that would see the latter buying “unlimited licences” for a fixed price.

“This way the government could save millions. It would not need to buy expensive foreign products and in addition would be buying Indian language products,” Annadurai argues.

Chennai Kavigal also plans to make inroads into the educational sector. “Unless we get into rural India, which has a huge but untapped market, attempts like this won’t survive,” he argues.

IIT collaboration
On the GNU/Linux front, Chennai Kavigal is working with the Indian Institute of Technology (IIT), Madras.

“There is a fantastic pool of talent there, especially as far as GNU/Linux is concerned,” says Annadurai. He points out that this team has been successful in making the “console level” of GNU/Linux work efficiently with Indian-languages. Currently, this solution works for software like Pine, GCC and Telnet meaning that anyone using these could now avail of an Indian-language interface.

“Now, they’re working on the kernel level, so that any application working on Linux will inherit Indian-language support,” says he. “Forget about the ISCII-versus-Unicode debate. Any solution should be encoding-free. Applications should be isolated from their code [to enable a suitable solution],” says Manoj with pragmatic finality