Why AI Agents Need Context | Deep Dives with a16z
a16z Deep Dives · 2026-06-02 · 50 min
https://www.youtube.com/watch?v=fuPgo7YcE7YVideo summary
Fivetran CEO George Fraser says AI agents need centralized business data, while warning AI-native rivals could outpace SaaS incumbents.
Fivetran CEO George Fraser tells Martin Casado that AI agents make centralized, continuously updated business data more important: without context, business AI resembles ChatGPT before internet access. He criticizes vendors that restrict access, citing SAP’s agent-access policy, and urges CIOs to keep a company-controlled data lake and negotiate data-access guarantees into major contracts. Fraser disputes data gravity, saying Fivetran’s change-data-capture approach moves only small volumes despite serving 7,000 large customers. He sees SaaS disruption less as wholesale replacement than a risk from AI-native competitors, while coding agents could improve Fivetran’s 750 connectors. OpenAI and Anthropic use conventional data platforms, he says, and Snowflake, Databricks, BigQuery, or Iceberg can support AI. Fivetran’s dbt merger combines data ingestion with modeling; Fraser’s main concern is DIY connectors, but he sees AI creating greater demand for data foundations and tools.
Chapters
- 0:00AI Agents Need Context: Fivetran Unifies Salesforce and NetSuite Data
- 3:36SaaS Lockdown Backlash: SAP Restricts Agent Access as Fraser Defends Open APIs
- 10:21Open Data Playbook: Open Data Infrastructure Flags SAP and Slack Restrictions
- 15:24Open Data Playbook: Fivetran’s Change Data Capture and Data-Access MSA Terms
- 19:47Enterprise Agents in Practice: Fivetran Uses Support and Salesforce CLI Agents
- 27:06Minimal Agents vs. MCP: Nano Claw’s Customization Trade-Off
- 29:01SaaSpocalypse and AI-Native Threats: Fivetran applies AI to 750 connectors
- 35:07AI Boosts Data Foundations: OpenAI and Anthropic use centralized data lakes
- 39:26dbt Merger and CEO Outlook: Coding agents can expand dbt modeling
- 44:33dbt Merger and CEO Outlook: DIY connector risk and AI-driven data opportunity
This is a Tier 1 public summary
Whether the chapter key points, section summaries and mind map are public is up to the person who shared it. Want the full analysis?Submit one yourself.
More from this channel
To Regulate AI Effectively, Focus on How It’s Useda16z Deep DivesMartin Casado argues AI laws should target harmful uses, while regulatory uncertainty pushes US startups toward Chinese open-source models.
Mintlify and the Transition From Human Docs to Agent Infrastructurea16z Deep DivesMintlify grew from eight pivots and a two-day prototype into documentation infrastructure serving AI agents and 20 million monthly visitors.
Inferact: Building the Infrastructure That Runs Modern AIa16z Deep DivesInferact’s vLLM founders aim to build an open inference layer for models running across 400,000–500,000 GPUs.
How Palantir Scaled: Why the Best Software Is Built Backwardsa16z Deep DivesPalantir architect Akshay Krishnaswamy explains how field engineers turn customer pain into products and scale software backwards.
Temporal CEO on AI Agents & The Future of Software | Deep Dives with a16za16z Deep DivesTemporal CEO Samar Abbas says durable execution will underpin long-running AI agents, with cloud handling 150,000 actions per second.
Related analyses
Peter Steinberger: "Fun Is Velocity"Y CombinatorPeter Steinberger says OpenClaw hit 4.7 million weekly downloads, but 9,500 configuration options and security work nearly overwhelmed him.
EP154 - AI 真的解決了數學世紀難題嗎？Anthropic 員工開始警告末日科技浪OpenAI 宣稱 AI 破解 Navier-Stokes 難題，耗費上千萬美元引發署名爭議；陶哲軒警告 AI 可能削弱數學訓練。
Anthropic IPO at Risk, Meta’s Muse Pop, Token Prices Fall, Open Source Gains Share, Alignment FailsAll-In PodcastMeta Muse hit No. 1 with 3 million downloads as open AI models captured 80% of token use, putting Anthropic’s IPO at risk.
New Grad求職有多困難？矽谷大裁員、AI取代工程師恐慌下靠什麼進入OpenAI、Google、LinkedIn、Salesforce哈佛姐夢遊矽谷 AliceInSiliconWonderlandBerkeley New Grad 求職率約三成，四位來賓分享靠 GitHub 職缺、校友推薦與真人模擬面試進入 OpenAI、Google 等公司。
Braintrust CEO on Where Engineering Actually Matters in AIa16z Deep DivesBraintrust CEO Ankur Goyal says AI engineering should focus on evals and harnesses, while SQL beat Bash in his agent benchmark.