Watch on Youtube.
On this session, UC At this time’s Kieran Devlin sits down with Kalyan Kumar (KK), Chief Product Officer at HCL Software program, to diagnose a essential challenge going through the World 2000: the lack to maneuver AI from the lab to the true world. With so many corporations caught operating limitless experiments with out delivering laborious enterprise outcomes, this dialog presents the architectural blueprint wanted to interrupt by way of the impasse. KK shares why the key to AI success isn’t really concerning the AI itself—it’s about the way you handle the info that feeds it.
Everyone seems to be dashing to roll out AI, however few are seeing the productiveness good points promised. Why? In keeping with KK, it’s not an AI drawback—it’s a knowledge drawback. The enterprise panorama is a “tangled internet” of disparate functions, and and not using a data-first working mannequin, deploying autonomous brokers typically ends in merely making unhealthy choices sooner.
On this deep dive, we discover why modernization doesn’t imply ripping out “traditional” programs like mainframes, however reasonably constructing an orchestration layer that connects them to new intelligence. KK explains why the long run isn’t nearly choosing an LLM, however about mastering metadata and getting ready for a multi-agent world the place governance is non-negotiable.
Key dialogue factors:
The Knowledge-First Crucial: Why you could untether knowledge from functions and grasp metadata earlier than AI can succeed—treating your enterprise like a well-organized library reasonably than a chaotic storage room.
Fixing the Integration Paradox: Find out how to bridge fashionable AI brokers with “traditional” core programs (ERPs, Mainframes) utilizing common orchestration reasonably than forcing a complete rip-and-replace modernization.
Governance in a Multi-Agent World: Making ready for the rise of Agent-to-Agent (A2A) communication and the Mannequin Context Protocol (MCP) to forestall autonomous brokers from creating battle.
