{"rewrite":{"id":"r_447033c8d8d4f7a23a0f4068","clusterId":"c_408af08f0fe45018cfeac980","slug":"amazon-cuts-ticket-classification-costs-98-with-strands-decider","model":"deepseek-v4-1-flash","headline":"Amazon Cuts Ticket Classification Costs 98% With Strands Decider","summary":"Amazon replaced Claude Sonnet 4.6 with its own decision-making model, Strands Decider, for sorting internal tickets into 20 categories, and cut processing costs by 98%. The tuned Strands Decider E4B, a 4-billion-parameter model based on Gemma 4, handles one million tickets for $500 instead of $22,200 and shortened average Lambda execution time by 74%.","whyItMatters":"Classification work that only needs a label, not generated text, can be done far more cheaply by a small decision model than by a large language model, and Amazon's own numbers put the gap at roughly 44 times the cost.","webCardHtml":"\u003cp\u003eAmazon\u0026#39;s internal ticket system sorts customer escalations, capacity requests, automated alarms, and security issues into 20 categories. Claude Sonnet 4.6 handled that classification until Amazon decided a model that outputs a probability and confidence score per option fit the job better than a large language model built for text and image generation.\u003c/p\u003e\u003cp\u003eThe tuned Strands Decider E4B returns only the option, not JSON, which shortens prompts and drops output tokens per ticket from 170 to 1. Input tokens fall from 6554 to 2009. Strands Agents cautions that a decision model may not replace an LLM when labels need explanatory text, the catalog changes weekly, or useful context cannot fit in a few thousand tokens.\u003c/p\u003e","blueskyPost":"Strands Decider E4B is a 4-billion-parameter model doing ticket sorting at $500 per million. The 98% saving comes from matching model size to task, not from a better model.","twitterPost":"Strands Decider E4B is 4 billion parameters. Amazon's 98% saving is a task-fit story, not a capability one.","threadsPost":"Strands Decider E4B runs 4 billion parameters on Gemma 4 and sorts tickets into 20 fixed categories. The 98% cut is a task-fit result: a small tuned model for narrow routing, not a general model replaced by a smarter one.","newsletterBlurb":"Amazon moved internal ticket classification off Claude Sonnet 4.6 and onto its own Strands Decider E4B, a 4-billion-parameter model based on Gemma 4. Cost for one million tickets fell from $22,200 to $500. Strands Agents notes the approach fits label-only work, not cases where labels need explanatory text or the catalog changes weekly.","attributionJson":"[{\"source\":\"GIGAZINE\",\"url\":\"https://gigazine.net/news/20261111-amazon-claude-sonnet-4-6-strands-decider/\",\"title\":\"Amazon succeeds in cutting costs by 98% by replacing AI model \\\"Claude Sonnet 4.6\\\" with its own decision-making model \\\"Strands Decider\\\"\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":5615,"outputTokens":773,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1794401294,"createdAt":"2026-11-11T12:45:32.000Z","publishedAt":"2026-11-11T12:47:11.000Z","updatedAt":"2026-11-11T12:47:11.000Z"},"cluster":{"id":"c_408af08f0fe45018cfeac980","canonicalTitle":"AmazonがAIモデルの「Claude Sonnet 4.6」を独自開発の意志決定モデル「Strands Decider」に置き換えることで98％のコスト削減に成功","representativeArticleId":"a_f7fb13a53245c0e3f5416448","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[],\"studios\":[],\"people\":[],\"type\":\"business\",\"domain\":\"industry\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-11-11T12:00:00.000Z","lastSeenAt":"2026-11-11T12:00:00.000Z","updatedAt":"2026-11-11T12:47:11.000Z"},"attribution":[{"source":"GIGAZINE","url":"https://gigazine.net/news/20261111-amazon-claude-sonnet-4-6-strands-decider/","title":"AmazonがAIモデルの「Claude Sonnet 4.6」を独自開発の意志決定モデル「Strands Decider」に置き換えることで98％のコスト削減に成功"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":[],"studios":[],"people":[],"type":"business","domain":"industry","is_roundup":false},"keyFacts":["Amazon replaced Claude Sonnet 4.6 with Strands Decider for classifying internal tickets into 20 categories.","Strands Decider E4B is a 4-billion-parameter model based on Gemma 4.","Input tokens per ticket fell from 6554 to 2009 and output tokens from 170 to 1.","Cost to process one million tickets dropped from $22,200 to $500, a 98% reduction.","Average Lambda execution time fell 74% after the migration to Strands Decider E4B."],"pageLinks":{"healthSlugs":[],"storySlugs":[],"entities":[],"stories":[],"entitiesOk":true,"storiesOk":true}}