Large langauge models developed in India and trained on data sets in vernacular languages are crucial for the development of AI in business applications from a language perspective, said Thomas Saueressig, a member of the executive board of German software giant SAP SE.
“1.4 billion people live in India, so Indian data and Indian large language models are going to be absolutely critical,” SAP SE’s deputy chairperson of the supervisory board Punit Renjen emphasized, when asked about the importance of Indian LLMs.
So far, two Indian large language models have been launched. Ola co-founder Bhavish Aggarwal unveiled a ‘Made in India’ large language model and generative AI platform, Krutrim, on the lines of OpenAI’s ChatGPT earlier this month. The platform is trained on 2 trillion tokens or pieces of textual information, having the largest representation of Indian data. Besides, Peak XV-backed Indian startup Sarvam has also launched the first Hindi large language model “OpenHathi”.
In a media briefing, the company also said that it has hired around 1,500 people this year for SAP Labs India and expects to hire a similar number of people next year. SAP Labs currently has about 15,000 employees in India, which the company plans to double in the coming few years.
“Nearly 40% of the global R&D workforce is based in India,” noted Sindhu Gangadharan, senior vice president and managing director at SAP Labs India. The company is building a new campus in Devanahalli, near Kempegowda International Airport, Bengaluru, which will house the new set of 15,000 employees when completed.
“The new campus will be operational in the beginning of 2025 and in the first phase we will have a capacity of 3,000 employees,” she added.
SAP’s management also emphasized on the growth opportunity it sees in India, with Renjen adding “This is India’s century.” The company is bullish on India as a marketplace and also as a talent pool for its research and development functions.