Watch on Youtube.
On this session, UC Immediately’s Kieran Devlin sits down with Kalyan Kumar (KK), Chief Product Officer at HCL Software program, to diagnose a vital problem going through the World 2000: the shortcoming to maneuver AI from the lab to the actual world. With so many corporations caught operating infinite experiments with out delivering laborious enterprise outcomes, this dialog gives 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 speeding to roll out AI, however few are seeing the productiveness positive factors promised. Why? Based on KK, it’s not an AI downside—it’s a knowledge downside. The enterprise panorama is a “tangled net” of disparate functions, and and not using a data-first working mannequin, deploying autonomous brokers usually leads to merely making unhealthy choices sooner.
On this deep dive, we discover why modernization doesn’t imply ripping out “basic” methods like mainframes, however fairly constructing an orchestration layer that connects them to new intelligence. KK explains why the longer term isn’t nearly selecting 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 it’s essential to untether information from functions and grasp metadata earlier than AI can succeed—treating your enterprise like a well-organized library fairly than a chaotic storage room.
Fixing the Integration Paradox: The right way to bridge fashionable AI brokers with “basic” core methods (ERPs, Mainframes) utilizing common orchestration fairly 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.
