L&T Technology Services, one of the leading global pure-play engineering services companies, is doubling down on generative AI. Making major announcements with giants like NVIDIA, Intel, Microsoft, Google, and AWS among others, LTTS is all set to accelerate the growth of AI startups and enterprise adoption in India.
In an exclusive interview with AIM, Abhishek Sinha, chief operating officer and board member at LTTS, said that over the past year, LTTS has trained over 3,000 engineers on generative AI technologies, surpassing its initial target of 2,000.
He also said that the company has developed a training program on generative AI for its top 200-300 leaders to enable them to have informed conversations with customers about the technology’s potential applications.
Meanwhile, top Indian IT companies like TCS, Infosys, and Wipro have trained about 825,000 employees combined in generative AI. A majority of Indian IT companies are now working with customers in deploying generative AI, going beyond PoCs.
LTTS told AIM that it is currently working on nine use cases in areas like digital twins for flexible production, AI-powered optical inspection, predictive maintenance, demand forecasting, supply chain management, and medical applications.
“As we started talking to customers, the customers started reaching out to say, ‘We are also exploring and are not sure, can you do a PoC for me?’” said Sinha.
LTTS is currently conducting over 90 PoCs for customers, some paid and some unpaid, demonstrating strong interest across industry segments.
LTTS has also filed more than 52 patents related to generative AI technologies in the past nine months. In the coming months, the company looks to build more use cases and train an additional 1,800 employees.
Recently, the company concluded a 6-8 week generative AI hackathon for employees, where they were given training and problem statements from the business. The hackathon was supported by hyperscalers like Google, Amazon and Microsoft.
GenAI Needs Time to Mature for Enterprises
While generative AI presents immense potential, Sinha acknowledged customers’ concerns around intellectual property and data security when engaging service providers. He expects the industry to evolve with the creation of ‘private clouds’ by customers to share non-core data with partners for generative AI applications.
For instance, Google is working with companies like Dataiku, Redis, SingleStore, and Starburst to integrate with its Vertex AI platform, enabling customers to train AI models using their data stored in Google Cloud.
Similarly, Snowflake and NVIDIA are collaborating to allow businesses to securely build and deploy custom AI models within the Snowflake Data Cloud using their own data.
“From a consumer perspective, generative AI is an amazing thing that has happened in the world,” said Sinha. He also believes that this technology has the power to revolutionise the way individuals interact with and benefit from AI-driven solutions.
However, when it comes to the business perspective, Sinha urged for a more cautious approach. “I think the time has not come yet,” he remarked, suggesting that the adoption of generative AI in the corporate world may require more time to mature.
Sinha acknowledged that product organisations, such as Google, are actively developing and releasing generative AI models.
“Now, within the product organisations like Google and others, they are creating these models. They’re putting it out there,” he said, recognising the efforts of these companies in pushing the boundaries of generative AI.
Sinha believes that investing in generative AI makes perfect business sense for companies like Google, highlighting the potential benefits and opportunities that generative AI presents for these organisations.
But for service providers, not so much. “I think there’s still some time to go,” he said, indicating that the evolution of generative AI in these contexts is still in its early stages.
“Customers are dabbling with the usage of GenAI, but they are doing so within their boundaries, within their companies. Their IP belongs to them. Their employees can do whatever play with that IP.
“So if you create a private cloud, private co-pilot or whatever for the customer data to be used by their employees, there’s no risk,” he remarked.
However, Sinha pointed out that exposing data beyond a company’s employees to outside parties introduces risks. He believes the industry will evolve to create secure clouds and bubbles, differentiating between core and non-core data.
Core data, such as critical algorithms, will remain protected, while non-core data may be shared with trusted partners.
In a recent conversation with AIM, Srinivas Konidena, CTO and VP at ADP, a payroll and HR system provider, echoed similar views and emphasised the importance of data security and privacy in their business.
“We are very paranoid about our data,” said Konidena, adding that “Whenever we make a decision, we are reminded not to think of it as our data. It’s the client’s and employee’s data.”
However, this is not the case with companies in the West; they are already at it in terms of implementing generative AI solutions. For instance, Accenture recorded a cumulative GenAI revenue of $1.1 billion during the first two quarters.
Capgemini claimed to be working on 300+ generative AI projects. Cognizant mentioned having 250 early engagements using its generative AI services and solutions, with 350+ opportunities in its pipeline.
