{"rewrite":{"id":"r_482d850f9b627f4914db6399","clusterId":"c_d1fc5018064b64103b6231e5","slug":"microsoft-research-unveils-memora-a-long-term-memory-architecture-for-ai-agents","model":"deepseek-v4-flash","headline":"Microsoft Research Unveils Memora, a Long-Term Memory Architecture for AI Agents","summary":"Microsoft Research released Memora, a long-term memory architecture for AI agents that balances storing fine details with efficient retrieval. It separates stored content from search entry points, using primary abstractions and cue anchors to allow agents to recall information from multiple clues, and includes a policy-guided retriever that revises search strategies when initial results are insufficient.","whyItMatters":"Memora addresses a core bottleneck for AI agents working over long periods by enabling them to retain detailed context without overwhelming token counts or losing specifics in summaries.","webCardHtml":"\u003cp\u003eMicrosoft Research published Memora on June 29, 2026, a memory architecture designed to let AI agents store conversation and task histories over long periods and retrieve only what is needed. The system separates stored content, called memory values, from search entry points: primary abstractions summarize the subject, while cue anchors provide alternative paths via related names, schedules, or topics. A policy-guided retriever iteratively refines the search strategy, expanding clues when information is lacking and stopping once enough is gathered. Memora was tested on the LoCoMo benchmark for long conversations.\u003c/p\u003e","blueskyPost":"Microsoft Research released Memora, a long-term memory architecture for AI agents that balances fine detail with easy retrieval. It separates stored content from search entry points and uses a policy-guided retriever to refine searches. Tested on the LoCoMo benchmark.","twitterPost":"Microsoft Research released Memora, a long-term memory architecture for AI agents. It stores detailed content separately from search clues, using primary abstractions and cue anchors. A policy-guided retriever revises search strategies. Tested on the LoCoMo benchmark.","threadsPost":null,"newsletterBlurb":"Microsoft Research released Memora, a long-term memory architecture for AI agents that balances preserving fine details with efficient retrieval. It separates stored content from search entry points and includes a policy-guided retriever that revises search strategies when initial results are insufficient.","attributionJson":"[{\"source\":\"GIGAZINE\",\"url\":\"https://gigazine.net/news/20260630-microsoft-memora-harmonic-memory/\",\"title\":\"What is Microsoft's 'Memora' Long-Term Memory Architecture That Prevents AI Agents from 'Forgetting'?\"}]","lintFlagsJson":null,"lintHits":0,"costUsd":0,"inputTokens":4160,"outputTokens":520,"status":"published","repairAttempts":0,"nextRepairAt":null,"factsAttemptedAt":1782808853,"createdAt":"2026-06-30T08:36:52.000Z","publishedAt":"2026-06-30T08:39:50.000Z","updatedAt":"2026-06-30T08:39:50.000Z"},"cluster":{"id":"c_d1fc5018064b64103b6231e5","canonicalTitle":"AIエージェントの「物忘れ」を防ぐMicrosoftの長期記憶アーキテクチャ「Memora」とは？","representativeArticleId":"a_8b3dfbbaae3d830100471029","sourceCount":1,"writtenSourceCount":1,"writeAttempts":0,"isSolo":true,"entitiesJson":"{\"anime_titles\":[],\"manga_titles\":[],\"work_titles\":[],\"studios\":[],\"people\":[],\"type\":\"news\",\"domain\":\"other\",\"is_roundup\":false}","contentType":"news","status":"published","firstSeenAt":"2026-06-30T07:55:00.000Z","lastSeenAt":"2026-06-30T07:55:00.000Z","updatedAt":"2026-06-30T08:39:50.000Z"},"attribution":[{"source":"GIGAZINE","url":"https://gigazine.net/news/20260630-microsoft-memora-harmonic-memory/","title":"AIエージェントの「物忘れ」を防ぐMicrosoftの長期記憶アーキテクチャ「Memora」とは？"}],"entities":{"anime_titles":[],"manga_titles":[],"work_titles":[],"studios":[],"people":[],"type":"news","domain":"other","is_roundup":false},"keyFacts":["Microsoft Research published Memora on June 29, 2026.","Memora separates stored content, called memory values, from search entry points: primary abstractions and cue anchors.","A policy-guided retriever iteratively refines search strategies, expanding clues when information is lacking and stopping once enough is gathered.","Memora was tested on the LoCoMo benchmark for long conversations."]}