Information Technology – Sector Update – From pilots to progress, enterprises build base for next phase of AI scaling

We attended multiple public forums and NASSCOM IT Conference to understand the ground pulse and progress around AI. The front-end business-centric use cases and pilots see notable conversions from experimentation to production, while the mid and back-office operations do not see relevant outcomes. Although the foundational models are in place with robust AI infrastructure, engineering and designing AI workloads into an Agentic form is a real challenge for enterprise clients. Using relevant data, establishing guardrails and introducing proper governance are some of the complexities that need to be addressed before AI agents take their decisions independently. More importantly, sourcing the responsible body, in case the agents compromise systems, is another challenge beyond the scope of these complexities. For IT Service providers, the deflationary impact of Advanced AI is yet to compensate against the proportionate surge in demand and appetite from enterprises clients that is long due. Probably the macro uncertainties and geo-political actions are weighing on their sentiment, or they have become more ROI-centric. The build initiatives on AI come at a higher pricing, but the overall pie itself is small to notably drive the needle, while disproportionate focus is delved into consolidation activities and discount asks. As the industry progresses through AI-led innovation, the pricing models from input-based to outcome-oriented would drive more relevance to enterprise clients and provide higher degree of control to outsourcing vendors. We prefer names, HCLT, INFY, PSYS and MPHL, that have demonstrated an ability to be an early mover during changing business narrative and have been frontrunners in capturing enterprise spending. In Small Cap, we prefer Latent View.