How Exa is Building the Perfect Search Engine | Deep Dives with a16z
a16z Deep Dives · 2026-06-04 · 49 min
https://www.youtube.com/watch?v=k6OkMkDdbUYVideo summary
Exa CEO Will Bryk argues agentic search will surpass Google by 2030 through deeper retrieval, comprehensive results, and AI-native infrastructure.
Exa cofounder and CEO Will Bryk explains why search built for AI agents must go far beyond Google’s human-click model. Agents need complex semantic and keyword queries, controllable filters, fresh data, and comprehensive outputs—potentially 1,000 to 10,000 results instead of 10. Bryk says Exa’s LLM-based retrieval can reduce customer costs by up to 20x, while reinforcement learning with Exa required fewer calls and delivered higher performance than traditional SERP tools. He predicts agentic search could exceed Google Ads in market size by 2030, driven by thousands or millions of searches from personal assistants and software agents. The conversation also covers coding agents, retrieval’s role in reducing token use, infrastructure and data bottlenecks, and Exa’s research-driven culture of hiring passionate, highly autonomous employees.
Chapters
- 0:00Exa Origin Story: From a College Search Engine to a 2021 Mission
- 2:35Beating Google With LLMs: Deep Queries, $10 Billion Costs, and ChatGPT
- 9:00Search Built for Agents: Complex Queries, Controls, and Complete Results
- 12:11Search Built for Agents: 10,000 Results and an Easier LLM-Native Architecture
- 15:47Search Built for Agents: 99.9999% Reliability and the Non-Commodity Frontier
- 19:03Search Built for Agents: Go-to-Market Intelligence, Polarization, and Retrieval
- 22:30Search Built For Agents: Exa Makes Coding Agents More Accurate
- 26:01Tokenpocalypse and Retrieval: Exa Claims 20x Cost Savings
- 29:21RL for Better Search: Exa Outperforms SERP in Agent Training
- 32:23Benchmarks and Agentic TAM: Agentic Search Could Exceed Google Ads by 2030
- 36:52Benchmarks and Agentic TAM: Infrastructure and Data Become the Bottlenecks
- 39:58Benchmarks and Agentic TAM: SpaceX Lessons, Exa’s Goat, and 10^18 Information
- 44:20Culture Hiring and Wrap: Exa’s Passion-Driven Culture and Agentic Hiring
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
Braintrust CEO on Where Engineering Actually Matters in AIa16z Deep DivesBraintrust’s Ankur Goyal contrasts AI’s Bitter Lesson with engineering, citing GLM5, SQL benchmarks, and Chinese models.
To Regulate AI Effectively, Focus on How It’s Useda16z Deep DivesMartin Casado argues AI laws should target harmful use, while US regulatory uncertainty pushes startups toward Chinese open-source models.
Inferact: Building the Infrastructure That Runs Modern AIa16z Deep DivesInferact founders explain how vLLM scaled from OPT to 500,000 GPUs and a universal inference layer
How Palantir Scaled: Why the Best Software Is Built Backwardsa16z Deep DivesPalantir’s FDE model builds Gotham and Foundry backward from mission outcomes, turning customer pain into reusable products.
Mintlify and the Transition From Human Docs to Agent Infrastructurea16z Deep DivesMintlify evolved through eight pivots into AI infrastructure, powering docs for agents, enterprises, and 20 million monthly readers.
Related analyses
2026/8/19(三)債市風暴 撿便宜時刻?美中爭霸燒錢戰 Al時代只能二選一？【早晨財經速解讀】游庭皓的財經皓角美債殖利率飆升，AI融資、伊朗戰事與中國模型掀起科技股估值回檔
How AI Will Rewire Consumer Internet | Deep Dives with a16za16z Deep DivesJosh Elman says ChatGPT, Siri, Robinhood, and TikTok show AI consumer products win through retention and better lives.
EP313. 谷歌痛失 AI 大將、Fable 5 配方解密、Meta 士氣低落中、FSD 即將入台 | M觀點M觀點Google痛失AI大將，Fable 5靠Agent Harness變強，Meta士氣崩盤、特斯拉FSD送審台灣
EP326. PLTR 怪物級財報、SPCX 的未來大餅、谷歌 AI 組織調整 | M觀點M觀點Palantir營收增92.8%、利潤率47.1%，SPCX押注AI與Starlink，Google重整DeepMind
EP327. 谷歌會放棄模型嗎、我低估了 SPCX、美股來到歷史新高 | M觀點M觀點Google不會放棄前沿模型，SpaceX估值上調至1.4兆美元