HCLTech is exploring small language models (SLMs) for select verticals like telecom and manufacturing Vijay Guntur, chief technology officer told Padmini Dhruvaraj. Internally, AI-powered tools have enhanced hiring, document management, and software delivery, achieving up to 40% efficiency gains for the company. Further, AI adoption is delivering strong ROI, particularly in healthcare and drug discovery. HCLTech sees a 5x data revenue opportunity from AI, with strategic investments in data platforms and metadata. Excerpts:
Q: What’s the idea behind building small language modules?
A: We are considering vertical-specific SLMs, but they are just one part of our focus on vertical solutions. We don’t need SLMs for everything. Some AI/ML solutions don’t require model changes and can use off-the-shelf models. We aim to build repeatable solutions—develop once, deploy across customers. We haven’t finalised yet on which verticals we plan to build SLMs, but telecom and manufacturing—especially plant-side—are strong possibilities. In utilities and dispersed spaces where surveillance is difficult, we see potential for drone- and video-based surveillance systems.
Q: How is HCLTech using AI and Generative AI internally to boost efficiencies, and have you observed tangible improvements?
A: Internal efficiencies have multiple aspects. At the core is a simple co-pilot—implementing co-pilots to improve productivity in enabling functions, sales, marketing, and delivery. Beyond that, we are building specific solutions for document management, knowledge management, pre-sales, and solution processes, actively deploying and using them. We are also improving HR processes with an agentic workflow integrated into our hiring system, significantly boosting efficiency and productivity. Over the last 6–9 months, this workflow has enhanced hiring conversion rates and the quality of talent, layering on top of existing systems.
Q: What specific demands are clients making regarding AI, automation, and efficiency gains? Do you see higher revenue potential from AI-based solutions or SLMs/LLMs?
A: Clients seek efficiency gains across IT operations, software development, and testing. They are open to sharing in managed services gains, and our AI Force has already been deployed in 30+ customers with positive results. SLMs and LLMs are just one part of the value chain. ROI considerations include model, infrastructure, and GPU costs. We take a full-stack approach rather than focusing on sub-components to maximise value for customers.
Q: How are clients realising a return on investment from their AI adoption?
A: There is significant ROI, especially in healthcare. We have seen many cases where ROI is evident. An example is clinician time, which is at a premium. Even saving less than five minutes a week per clinician translates to $100 million in savings for a company with 20,000 clinicians. Building this system cost around $10 million.
In life sciences, drug discovery is risky and time-consuming. AI helps streamline this process, increasing efficiency and profitability. Beyond financial ROI, there’s a broader impact on people and society.
Q: What new initiatives or investments are underway in your ER&D labs, particularly in data and AI engineering?
A: While the automotive sector has been slow for us, we are investing in data, AI systems, and AI engineering—our third offering, focused on embedding AI into products. We have also developed multi-modal capabilities in video, audio, and AI-driven solutions to accelerate new product development.
