How agent o11y differs from traditional o11y
summary
Agent traces routinely exceed a gigabyte and single spans hit 20 megabytes, containing vast semistructured and unstructured text that must be ingested in real time. Braintrust builds a custom database with a write ahead log for instant visibility, analytical indexes for fast filtering, and a forked Tantivy full‑text index so engineers can query every trace that mentions a given word. Clinicians, lawyers, and wealth advisers open traces to grade responses, their written justifications become training signals, and automated scoring functions scale those failure modes.
AI-written summary · about
chapters (6)
- 0:00Introduction
- 2:41Traditional observability
- 5:09Agent o11y differences
- 8:59Agent traces complexity
- 12:08Agent o11y users
- 13:56Future of agent o11y & Q&A
related talks
25mFrom Blind Spots to Total Vision: O11y at Massive Scale
45mNever Send a Human to do an Agent's Search
31mDeep Dive To AI Agent Observability
42mAgents of Change: From Time Saved to Intelligence Activated
27mWhy Are Your AI’s Decisions Hard To Explain: Trace Every Decis...
45mAI agents in practice - beyond the prototype
47mAgents and Memory Architecture
43mArchitectural Blueprints for AI Agents