AI Agents and the Fight for Customer Data (51 min)

← Back to week overview

• Release date: 2026-06-02
• Listen on Spotify: Open episode
• Episode description:

  Martin Casado speaks with George Fraser, cofounder and CEO of Fivetran, about the future of data infrastructure in the age of AI. The conversation covers Fivetran’s merger with dbt, the changing role of data platforms, and why Fraser believes many companies are overestimating the threat AI poses to enterprise software. They discuss open data access, the backlash against AI agents accessing systems of record, and why businesses still need centralized data foundations even as agent-based workflows become more common. Along the way, Fraser shares his views on data gravity, coding agents, enterprise AI adoption, and how AI is changing the way software companies build and operate products.   Resources Follow George Fraser on X: https://x.com/frasergeorgew Follow Martin Casado on X: https://x.com/martin_casado Check out everything a16z is doing with artificial intelligence here, including articles, projects, and more podcasts. Please note that the content here is for informational purposes only; should NOT be taken as legal, business, tax, or investment advice or be used to evaluate any investment or security; and is not directed at any investors or potential investors in any a16z fund. a16z and its affiliates may maintain investments in the companies discussed. For more details please see a16z.com/disclosures. Hosted by Simplecast, an AdsWizz company. See pcm.adswizz.com for information about our collection and use of personal data for advertising.

Summary

• 📊 Data Centralization for Agents: AI agents need unified context from all SaaS and internal systems, extending the historical need for BI data platforms.
• 🔒 Vendor API Lockdowns: SaaS vendors are restricting agent access fearing disintermediation, but this harms customers and may be temporary.
• 👥 Agents as Team Members: Treating agents like employees with identities and workflows could increase rather than decrease software usage.
• 🚫 Data Gravity Myth: Egress fees and data movement costs are overstated; CDC makes replication far cheaper than assumed.
• 🛠️ Infrastructure Evolution: AI increases demand for robust data infrastructure rather than commoditizing it, with new opportunities in agent tooling.

Insights

• Should enterprises treat AI agents as additional human-like employees with their own identities, Slack access, and HR onboarding rather than pure software?
  • Time: 0:35 – 0:46
  • Answer: The conversation explores agents having separate emails, roles, and workflows to slot into existing human-designed systems, potentially increasing software consumption instead of reducing it. This intermediate approach avoids full system refactoring.
• Why do AI agents require centralized business data in ways that fundamentally change how companies should build their data infrastructure?
  • Time: 2:35 – 2:44
  • Answer: The transcript emphasizes that AI agents need broad context from all systems of record (Salesforce, SAP, etc.) to function effectively, similar to how BI reporting historically drove data centralization but now extended to agentic use cases. This shifts priorities from siloed SaaS tools to unified data platforms.
• Are SaaS companies overreacting by locking down APIs against AI agents, or is this a legitimate defense against disintermediation?
• Is the notion of data gravity overstated, and how does change data capture challenge assumptions about moving large enterprise datasets?

Want the full treatment? Download expanded insights—timestamps, complete answers, and course materials from AI Weekly Insights.