Watch on Youtube.
On this session, UC Right now’s Kieran Devlin sits down with Kalyan Kumar (KK), Chief Product Officer at HCL Software program, to diagnose a essential subject dealing with the World 2000: the lack to maneuver AI from the lab to the actual world. With so many companies caught working infinite experiments with out delivering onerous enterprise outcomes, this dialog affords the architectural blueprint wanted to interrupt via the impasse. KK shares why the key to AI success isn’t truly in regards to 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 beneficial properties promised. Why? Based on KK, it’s not an AI downside—it’s a knowledge downside. The enterprise panorama is a “tangled internet” of disparate purposes, and and not using a data-first working mannequin, deploying autonomous brokers typically leads to merely making dangerous selections quicker.
On this deep dive, we discover why modernization doesn’t imply ripping out “traditional” methods like mainframes, however quite 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 making ready for a multi-agent world the place governance is non-negotiable.
Key dialogue factors:
The Knowledge-First Crucial: Why you will need to untether knowledge from purposes and grasp metadata earlier than AI can succeed—treating your enterprise like a well-organized library quite than a chaotic storage room.
Fixing the Integration Paradox: How you can bridge fashionable AI brokers with “traditional” core methods (ERPs, Mainframes) utilizing common orchestration quite than forcing a complete rip-and-replace modernization.
Governance in a Multi-Agent World: Getting ready for the rise of Agent-to-Agent (A2A) communication and the Mannequin Context Protocol (MCP) to stop autonomous brokers from creating battle.
