Eps.80 | Agents never sleep: Teradata's Sumeet Arora on why the autonomous enterprise demands a new tech stack

Oct 28, 2025

Ask episode

Chapters

Transcript

Episode notes

Sumeet Arora, Chief Product Officer at Teradata, joins Tim Bond to discuss why the shift to autonomous enterprises demands far more than simply deploying AI agents. With Teradata serving major banks, retailers, and global organisations for over four decades, Sumeet explains how the traditional enterprise stack of hundreds of workflow-centric applications is evolving into knowledge fabrics that power agent layers with outcome-oriented interfaces.

The conversation explores Teradata's three competitive differentiators: efficient processing engines built for always-on agent workloads, true on-premises and multi-cloud capability that brings AI to where data lives, and industry knowledge models developed across 10 sectors that make agents enterprise-grade rather than experimental. Sumeet shares real examples, including SECI Brazil's 200% increase in credit offers without changing delinquency rates, and explains why 80% of AI projects are not moving from demos to deployment.

Timestamps

00:01:05 - Announcing Agent Builder: Teradata's platform for autonomous AI agents.
00:03:13 - The evolution to autonomous enterprise: from connected to digital to autonomous.
00:04:19 - How the enterprise tech stack is changing from applications to knowledge fabrics.
00:08:02 - Why starting with clear, measurable outcomes is essential for AI ROI.
00:11:04 - Foundation models: bring your own LLM or use Teradata's small, medium, large options.
00:12:02 - Agent builder layer: open-source integration with MCP server.
00:13:24 - User experience evolution: from workflow-oriented to outcome-oriented.
00:16:25 - Teradata's three competitive differentiators explained.
00:17:08 - Why agents never sleep: the efficiency advantage of always-on workloads.
00:18:00 - On-premises and multi-cloud: bringing AI to where your data actually lives.
00:20:17 - Industry knowledge models: data models across 10 industries developed over decades.
00:24:48 - Real customer examples: from credit risk to fraud reduction to compliance.
00:28:55 - Organisation readiness: training every employee to build and use agents.
00:31:13 - The people side: augmentation versus replacement in the AI era.
00:34:15 - Going back to basics: AI requires clean code, knowledge and data.
00:36:00 - The architecture conversation enterprises are not having yet.

Guest bio

Sumeet Arora is Chief Product Officer at Teradata, where he leads product strategy for one of the world's leading enterprise data and analytics platforms. With extensive experience at Cisco, building secure, affordable internet infrastructure and at ThoughtSpot, democratising data analytics for business users, Sumeet focuses on changing industries for the better. At Teradata, he is driving the evolution from data platforms to knowledge platforms that enable the autonomous enterprise.

Resources and links

Sumeet Arora on LinkedIn
Tim Bond on LinkedIn
Teradata Agent Builder information

This episode is brought to you by Cogniscale, which helps teams build and scale their AI workforce.

Want to teach your team to build their own AI agents that act like teammates, boosting productivity, quality, and job satisfaction? Get your FREE AI Acceleration Report (for listeners with teams of six or more), a quick team survey and consultation that maps priority workflows, time saved and ROI from an AI upskilling programme. Visit Cogniscale.com and enter promo code PODCAST.

If you enjoyed this episode, please follow Enterprise Thought Leadership on your preferred platform and rate the show. Have burning topics or questions you would like explored? Email the team at podcasts@techpros.io