Infosys: Infosys launches generative AI services suite

BENGALURU: Infosys has launched a suite of AI services using generative AI that it says will augment human capabilities and open new opportunities in the enterprise space.
Infosys Topaz, as it is called, leverages the company’s AI framework to help customers build an AI core to deliver cognitive solutions. “It’s our collective knowledge, services, and the ecosystem of partners that we work with that we will bring to bear for our clients. Our clients need a lot of advisory capability around it. One of the first things that we are doing which Nandan (Nilekani) himself is championing is an AI-first approach within the organisation, and we’re going to bring in those learnings into our Topaz offerings, because harnessing generative AI needs a lot of experience,” Infosys EVP and co-head of delivery Satish HC told TOI.
A recent Goldman Sachs report said generative AI could boost economic output by $7 trillion globally over the next ten years. US investment management firm Ark Invest said AI software could generate $14 trillion in revenue by 2030.
“Even without generative AI, the market was about $400-$500 million. This is a high-growth opportunity and AI is at the heart of everything that we do. We’re building digital twins for ourselves. One of the biggest challenges in the industry today, not just for us but even for our clients, is the aging workforce, institutional knowledge that is sitting within teams, and how we harness it without losing it. With generative AI now, we have an answer. Second, we are heavily invested in a core team,” Satish added.
Satish noted that for a customer, they had been using machine learning algorithms that can understand emotions and sentiments around their products during sporting events. “With ChatGPT, we’ve figured out that we can do this with 30% less effort. In some of these cases, we’re going to use ChatGPT because it’s going to work on data which is already there in the public domain in social media. Now, in many other cases, we will have to build a custom module, or a large database bordered within the confines of a company because it’s got competitive knowledge,” he said.
A British bank, he said, used Infosys Topaz to transform over 2,000 customer service processes to operate in near real-time instead of a week.
Infosys has over 12,000 AI use cases, 150 pre-trained AI models and 10 AI platforms steered by AI and data strategists. It says it has embraced a responsible by design approach that ensures trust, privacy, security, and compliance.
Satish said that Infosys has built a risk framework around it and captured generative AI projects in the master service agreement (MSA) to mitigate any downside impact. “Contractually, every place where we are experimenting with this technology, we only do this with the explicit concurrence and approval from our clients and it goes into the MSA. That’s because of the indirect risks which are open-ended. Second, when we take our service, whether it is 95% accuracy, whether it is 75% accuracy, the balance, we underwrite. Like the example I shared on 30% productivity gains with ChatGPT, it has given me a cycle time and effort savings. Of course, there may be a rough edge, but we will continue to fine-tune the models. And we will underwrite the rest of the accuracy gap, so that it is seamless for our clients,” he said.

Leave a Comment

Your email address will not be published. Required fields are marked *