--- title: Citation Needed: Provenance for LLM-Built Knowledge... description: A conference talk by Daniel Chalef of Zep AI covering the challenge of data provenance in LLM-built knowledge graphs. LLMs synthesize facts... canonical: https://daily.dev/posts/citation-needed-provenance-for-llm-built-knowledge-graphs-daniel-chalef-zep-ai-dg0yq9yru twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Citation Needed: Provenance for LLM-Built Knowledge Graphs — Daniel Chalef, Zep AI | daily.dev og:description: A conference talk by Daniel Chalef of Zep AI covering the challenge of data provenance in LLM-built knowledge graphs. LLMs synthesize facts... og:url: https://daily.dev/posts/citation-needed-provenance-for-llm-built-knowledge-graphs-daniel-chalef-zep-ai-dg0yq9yru og:image: https://api.daily.dev/og/posts/Dg0yQ9yru.png og:image:alt: Citation Needed: Provenance for LLM-Built Knowledge Graphs — Daniel Chalef, Zep AI og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Citation Needed: Provenance for LLM-Built Knowledge Graphs — Daniel Chalef, Zep AI **[AI Engineer](https://daily.dev/sources/aidotengineer)** · 20 min read · 1 upvotes · 1 comments ## Summary A conference talk by Daniel Chalef of Zep AI covering the challenge of data provenance in LLM-built knowledge graphs. LLMs synthesize facts non-deterministically from multiple sources, destroying the paper trail of how outputs were derived. The talk introduces Graphiti, an open-source temporal graph framework, which models provenance by linking derived facts back to their source episodes as graph relationships. Key topics include: handling entity merges while preserving source links, metadata tagging for veracity filtering, partial deletion under GDPR/right-to-be-forgotten policies, and the engineering pipeline for entity extraction, deduplication, and fact deconfliction. The core argument is that lineage must be built into the data structure itself rather than logged after the fact. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Community discussion Top comments from developers on daily.dev. **@agustinbarrientos** · 0 upvotes > After deleting one source, what happens to the identity and edges of an entity merged from several episodes? ## Similar posts on daily.dev - [Context Graphs & Agentic Decisions](https://daily.dev/posts/context-graphs-agentic-decisions-ve9azxujl) · neo4j · 1 upvotes · 0 comments - [Agent Memory Is Only as Good as Its Schema](https://daily.dev/posts/agent-memory-is-only-as-good-as-its-schema-vamsh2ow0) · Daily Dose of Data Science \| Avi Chawla \| Substack · 0 upvotes · 0 comments - [Why Your Agent Remembers Everything and Understands Nothing](https://daily.dev/posts/why-your-agent-remembers-everything-and-understands-nothing-igylroaz2) · Daily Dose of Data Science \| Avi Chawla \| Substack · 1 upvotes · 0 comments - [The Problem of Provenance: Nobody Knows What’s In the Training Data](https://daily.dev/posts/the-problem-of-provenance-nobody-knows-what-s-in-the-training-data-p0me2twe9) · Medium · 0 upvotes · 0 comments --- Tags: [#llm](https://daily.dev/tags/llm) [View this post on daily.dev](https://daily.dev/posts/citation-needed-provenance-for-llm-built-knowledge-graphs-daniel-chalef-zep-ai-dg0yq9yru) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Citation Needed: Provenance for LLM-Built Knowledge Graphs — Daniel Chalef, Zep AI","url":"https://daily.dev/posts/citation-needed-provenance-for-llm-built-knowledge-graphs-daniel-chalef-zep-ai-dg0yq9yru","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/citation-needed-provenance-for-llm-built-knowledge-graphs-daniel-chalef-zep-ai-dg0yq9yru"},"datePublished":"2026-07-23T04:22:45.993Z","dateModified":"2026-07-23T04:23:06.010Z","description":"A conference talk by Daniel Chalef of Zep AI covering the challenge of data provenance in LLM-built knowledge graphs. LLMs synthesize facts...","image":"https://i.ytimg.com/vi/H7puB0RwJMM/sddefault.jpg","thumbnailUrl":"https://i.ytimg.com/vi/H7puB0RwJMM/sddefault.jpg","isAccessibleForFree":true,"articleSection":"AI Engineer","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"AI Engineer","logo":"https://media.daily.dev/image/upload/s--u5PucxNT--/f_auto/v1724338940/logos/aidotengineer","url":"https://daily.dev/sources/aidotengineer"},"commentCount":1,"discussionUrl":"https://daily.dev/posts/citation-needed-provenance-for-llm-built-knowledge-graphs-daniel-chalef-zep-ai-dg0yq9yru","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":1}],"keywords":"llm","timeRequired":"PT20M","video":{"@type":"VideoObject","name":"Citation Needed: Provenance for LLM-Built Knowledge Graphs — Daniel Chalef, Zep AI","description":"A conference talk by Daniel Chalef of Zep AI covering the challenge of data provenance in LLM-built knowledge graphs. LLMs synthesize facts...","thumbnailUrl":"https://i.ytimg.com/vi/H7puB0RwJMM/sddefault.jpg","uploadDate":"2026-07-23T04:22:45.993Z","duration":"PT20M","url":"https://api.daily.dev/r/Dg0yQ9yru","embedUrl":"https://www.youtube.com/embed/H7puB0RwJMM"}} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"AI Engineer","item":"https://daily.dev/sources/aidotengineer"},{"@type":"ListItem","position":3,"name":"Citation Needed: Provenance for LLM-Built Knowledge Graphs — Daniel Chalef, Zep AI"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/citation-needed-provenance-for-llm-built-knowledge-graphs-daniel-chalef-zep-ai-dg0yq9yru","comment":[{"@type":"Comment","text":"After deleting one source, what happens to the identity and edges of an entity merged from several episodes?","datePublished":"2026-08-02T22:01:21.601Z","url":"https://daily.dev/posts/Dg0yQ9yru#c-haVp173Sz","author":{"@type":"Person","name":"Agustin Barrientos","url":"https://daily.dev/agustinbarrientos","image":"https://media.daily.dev/image/upload/s--5ayxQnqn--/f_auto/v1788281802/avatars/avatar_wQYYVe5Tbj0NJ7C7qPoa8?_a=BAMAMicg0"}}]} ```