--- schema: 1 type: video title: How we solved Context Management in Agents — Sally-Ann Delucia description: >- The naive solution is truncation. The obvious solution is summarization. Neither worked — and the Arize team found out the hard way while building an AI agent that had to analyze the very trace data it was generating. A year of lessons from building Alyx, starting with the vicious loop that defined the problem: Alex runs on trace data, the spans grow, the context limit hits, it fails and tries again. The talk covers why truncation breaks reasoning, why summarization gives the LLM too much control, and how head/tail preservation with a retrievable memory store is what actually held. Then: long session evals, sub-agents as the answer when one context accumulates too much, and what they found when they went looking for secrets in the Claude Code source release. Speaker info: - https://www.linkedin.com/in/sallyann-delucia-59a381172/ Timestamps: 0:00 Introduction and speaker background 1:02 Overview of the AI agent, Alyx 1:29 The problem: Context engineering vs. prompt engineering 4:06 The vicious loop of data growth in AI agents 5:16 Why naive truncation failed 6:14 Why summarization proved unreliable 6:46 The solution: Smart truncation and memory stores 8:02 Handling long session challenges 9:23 Offloading tasks to sub-agents 11:19 Ongoing challenges and future work 12:57 Findings from the Claude Code source release 13:44 Final key takeaways on context management 14:58 Q&A session --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/esY99nYXxR4-how-we-solved-context-management-in-agents-sally-ann-delucia Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- canonical: >- https://liquidgold.cc/channel/aidotengineer/explore/video/how-we-solved-context-management-in-agents-sally-an-49f9c3bc30cf markdown: >- https://liquidgold.cc/channel/aidotengineer/video/how-we-solved-context-management-in-agents-sally-an-49f9c3bc30cf.md llms: null parent: 'https://liquidgold.cc/channel/aidotengineer.md' channel: aidotengineer updated: '2026-10-24T14:11:29.188Z' language: null children: [] videoId: esY99nYXxR4 durationSeconds: 977 publishedAt: '2026-05-10T19:00:06.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 103 endSeconds: 246 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=103' heldReason: rights - ordinal: 1 text: null startSeconds: 103 endSeconds: 246 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=103' heldReason: rights - ordinal: 2 text: null startSeconds: 257 endSeconds: 324 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=257' heldReason: rights - ordinal: 3 text: null startSeconds: 335 endSeconds: 378 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=335' heldReason: rights - ordinal: 4 text: null startSeconds: 389 endSeconds: 412 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=389' heldReason: rights - ordinal: 5 text: null startSeconds: 412 endSeconds: 423 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=412' heldReason: rights - ordinal: 6 text: null startSeconds: 498 endSeconds: 565 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=498' heldReason: rights - ordinal: 7 text: null startSeconds: 575 endSeconds: 687 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=575' heldReason: rights - ordinal: 8 text: null startSeconds: 697 endSeconds: 786 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=697' heldReason: rights - ordinal: 9 text: null startSeconds: 786 endSeconds: 797 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=786' heldReason: rights - ordinal: 10 text: null startSeconds: 797 endSeconds: 832 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=797' heldReason: rights - ordinal: 11 text: null startSeconds: 918 endSeconds: 977 watch: 'https://liquidgold.cc/watch/esY99nYXxR4?t=918' heldReason: rights --- # How we solved Context Management in Agents — Sally-Ann Delucia The naive solution is truncation. The obvious solution is summarization. Neither worked — and the Arize team found out the hard way while building an AI agent that had to analyze the very trace data it was generating. A year of lessons from building Alyx, starting with the vicious loop that defined the problem: Alex runs on trace data, the spans grow, the context limit hits, it fails and tries again. The talk covers why truncation breaks reasoning, why summarization gives the LLM too much control, and how head/tail preservation with a retrievable memory store is what actually held. Then: long session evals, sub-agents as the answer when one context accumulates too much, and what they found when they went looking for secrets in the Claude Code source release. Speaker info: - https://www.linkedin.com/in/sallyann-delucia-59a381172/ Timestamps: 0:00 Introduction and speaker background 1:02 Overview of the AI agent, Alyx 1:29 The problem: Context engineering vs. prompt engineering 4:06 The vicious loop of data growth in AI agents 5:16 Why naive truncation failed 6:14 Why summarization proved unreliable 6:46 The solution: Smart truncation and memory stores 8:02 Handling long session challenges 9:23 Offloading tasks to sub-agents 11:19 Ongoing challenges and future work 12:57 Findings from the Claude Code source release 13:44 Final key takeaways on context management 14:58 Q\&A session --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/esY99nYXxR4-how-we-solved-context-management-in-agents-sally-ann-delucia Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- Canonical: https://liquidgold.cc/channel/aidotengineer/explore/video/how-we-solved-context-management-in-agents-sally-an-49f9c3bc30cf · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md) > Search chunk text is held until public-source provenance is verified. ## Document chunks > Missing: no stored document chunks are available.