Stenobird

Episode

Glean's Waldo: The Agentic Search Model Making AI Faster and Cheaper

Podcast
Answer Engine Optimization (AEO): The AI Search Podcast
Published
May 2, 2026
Duration seconds
394
Processing state
processed
Canonical source
https://share.transistor.fm/s/52bff747
Audio
https://media.transistor.fm/52bff747/413dfcbb.mp3
JSON
/v1/public/podcasts/answer-engine-optimization-aeo-the-ai-search-podcast-7756998/episodes/glean-s-waldo-the-agentic-search-model-making-ai-faster-and-cheaper
Markdown
/podcast/answer-engine-optimization-aeo-the-ai-search-podcast-7756998/glean-s-waldo-the-agentic-search-model-making-ai-faster-and-cheaper.md

Actions

  • POST https://stenobird.com/v1/public/podcasts/answer-engine-optimization-aeo-the-ai-search-podcast-7756998/episodes/glean-s-waldo-the-agentic-search-model-making-ai-faster-and-cheaper/transcription-requests
    Idempotently request low-priority transcript generation for this episode.
  • GET https://stenobird.com/podcast/answer-engine-optimization-aeo-the-ai-search-podcast-7756998/glean-s-waldo-the-agentic-search-model-making-ai-faster-and-cheaper.md
    Read the agent-friendly Markdown representation of this episode resource.

Summary

Glean's Waldo introduces an agentic search model that uses reinforcement learning to optimize the retrieval process. By acting as a specialized intermediary, it significantly reduces latency and token consumption without sacrificing final answer quality.

Topics

  • Agentic Search
  • Glean Waldo
  • Reinforcement Learning
  • Enterprise AI
  • LLM Efficiency
  • Answer Engine Optimization
  • Query Decomposition
  • Retrieval Augmented Generation

Highlights

  • Main idea: Waldo functions as a specialized agent that performs query decomposition and tool selection before hitting frontier models
  • Efficiency gain: The model achieves 50% lower latency and 25% fewer tokens consumed through optimized retrieval
  • Technical mechanism: The system uses a planning loop of query decomposition, iterative search, and evaluation rather than a single lookup
  • Strategic shift: The rise of agentic search necessitates a move from traditional SEO to Answer Engine Optimization (AEO) to ensure brand discoverability
  • Failure mode: Relying on monolithic models for complex enterprise queries leads to unsustainable costs and slow response times

Chapters

  1. 0:00 Introduction to Waldo: An introduction to Glean's new agentic search model and its impact on enterprise AI efficiency.
  2. 1:00 The Problem with Massive Models: Discussing the latency and token costs associated with sending every query directly to frontier LLMs.
  3. 2:00 Efficiency Without Quality Loss: How Waldo uses a curated set of evidence to maintain high-quality answers while reducing compute.
  4. 3:00 The Agentic Reasoning Loop: A deep dive into query decomposition, tool selection, and the iterative search process.
  5. 4:00 The Future of Enterprise AI: Analyzing the shift from monolithic models to orchestrated systems of specialized agents.
  6. 5:00 Optimizing for Agentic Discovery: How brands must adapt their content structure to be found by intermediate search agents.
  7. 6:00 Conclusion and Takeaways: Final thoughts on the reality of agentic search models and the changing landscape of AI information consumption.