Stenobird

Episode

#353 The Data Team's Agentic Future with Ketan Karkhanis, CEO at ThoughtSpot

Podcast
DataFramed
Published
Mar 30, 2026
Duration seconds
2986
Processing state
processed
Canonical source
https://www.datacamp.com/podcast
Audio
https://dts.podtrac.com/redirect.mp3/cohst.app/pdcst/6G1A6D/episodes.captivate.fm/episode/99b8dd9a-bda4-4a8a-979c-a317372a95aa.mp3
JSON
/v1/public/podcasts/dataframed/episodes/353-the-data-team-s-agentic-future-with-ketan-karkhanis-ceo-at-thoughtspot
Markdown
/podcast/dataframed/353-the-data-team-s-agentic-future-with-ketan-karkhanis-ceo-at-thoughtspot.md

Actions

  • POST https://stenobird.com/v1/public/podcasts/dataframed/episodes/353-the-data-team-s-agentic-future-with-ketan-karkhanis-ceo-at-thoughtspot/transcription-requests
    Idempotently request low-priority transcript generation for this episode.
  • GET https://stenobird.com/podcast/dataframed/353-the-data-team-s-agentic-future-with-ketan-karkhanis-ceo-at-thoughtspot.md
    Read the agent-friendly Markdown representation of this episode resource.

Summary

The bottleneck for AI-driven analytics is not model capability, but data readiness and semantic clarity. Ketan Karkhanis explains how data teams must transition from building dashboards to designing governed, agent-friendly metadata.

Topics

  • AI Agents
  • Data Strategy
  • Business Intelligence
  • Data Governance
  • Analytics Engineering
  • Semantic Layer
  • Data Culture
  • Machine Learning Operations

Highlights

  • Main idea: AI agents require 'agent-friendly' metadata and governed semantics rather than just human-readable descriptions
  • Practical takeaway: Focus on delivering measurable ROI within 30 days on a single use case to drive organizational adoption
  • Failure mode: Avoid 'AI slop' by ensuring data quality and preventing the pursuit of perfection from stalling progress
  • Strategic shift: Data analysts must evolve into 'AI stewards' who manage the logic and trust layers of the data ecosystem
  • Critical warning: Do not mistake simple text-to-SQL capabilities for true generative AI innovation

Chapters

  1. 1:00 The Imperative of Agent-Friendly Data: Why column descriptions and metadata must be optimized for machines to prevent 'AI slop'.
  2. 4:40 The Reality of Self-Service BI: Moving beyond the decade-long failed promise of self-service toward true agentic automation.
  3. 8:20 The Evolution of the Analyst Role: How the rise of agents shifts the data professional's job toward governance and design.
  4. 19:30 Modern Data Stack Integration: Leveraging existing cloud data warehouses like Snowflake, Databricks, and BigQuery for AI readiness.
  5. 26:50 Building an AI-First People Strategy: Why organizational change management and new skill investment are as vital as the technology itself.
  6. 30:30 Avoiding the 'Shiny Demo' Trap: Distinguishing between meaningful AI innovation and basic text-to-SQL automation.
  7. 38:20 Driving ROI through Change Management: How to tie AI projects to concrete business value and decommissioning legacy technical debt.