--- title: Airbnb Cuts Recommendation Staleness from 2 Days to... description: Airbnb extended its open source Chronon feature platform with Push Mode and Near Real Time Model Transform to refresh recommendation embeddings within seconds... canonical: https://daily.dev/posts/airbnb-cuts-recommendation-staleness-from-2-days-to-seconds-with-chronon-xwnl5ftip twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Airbnb Cuts Recommendation Staleness from 2 Days to Seconds with Chronon | daily.dev og:description: Airbnb extended its open source Chronon feature platform with Push Mode and Near Real Time Model Transform to refresh recommendation embeddings within seconds... og:url: https://daily.dev/posts/airbnb-cuts-recommendation-staleness-from-2-days-to-seconds-with-chronon-xwnl5ftip og:image: https://api.daily.dev/og/posts/xwnl5FTIp.png og:image:alt: Airbnb Cuts Recommendation Staleness from 2 Days to Seconds with Chronon 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. # Airbnb Cuts Recommendation Staleness from 2 Days to Seconds with Chronon **[InfoQ](https://daily.dev/sources/infoq)** · 3 min read · 0 upvotes · 0 comments ## Summary Airbnb extended its open source Chronon feature platform with Push Mode and Near Real Time Model Transform to refresh recommendation embeddings within seconds of guest activity. This cut embedding staleness from nearly two days to under a minute, with updates typically arriving in 10 to 30 seconds. Offline NDCG improved by 1.67%, and online A/B testing showed about a 0.33% increase in uncancelled bookings. Push Mode publishes a notification when a feature value is committed, and NRT Model Transform runs the existing Transformer sequence encoder inside that streaming pipeline instead of a separate batch job. Both capabilities have been contributed to the public Chronon repository. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### How did Airbnb reduce staleness in its recommendation embeddings using Chronon? Airbnb added Push Mode and Near Real-Time Model Transform to its Chronon feature platform, letting a streaming job publish a notification when a feature value changes and run the already-deployed Transformer sequence encoder inside that streaming pipeline. This cut embedding staleness from nearly two days to under a minute, typically 10 to 30 seconds. _Developers designing low-latency feature pipelines can follow architecture breakdowns like this one on daily.dev._ ### What business impact did Airbnb see from switching to near real-time embeddings in Chronon? Airbnb measured a 1.67% improvement in offline NDCG and an approximately 0.33% increase in uncancelled bookings during online A/B testing after deploying near real-time embedding updates through Chronon's Push Mode and NRT Model Transform. _Teams justifying real-time ML infrastructure investments can track results like these on daily.dev._ ### What is NRT Model Transform in Airbnb's Chronon feature platform? NRT Model Transform is a Chronon capability that lets an already deployed model run directly inside a streaming pipeline, so the same Transformer sequence encoder used for daily batch processing can also generate embeddings in near real time, avoiding a separate inference implementation for the streaming workflow. _Engineers evaluating streaming inference architectures can follow platform updates like this on daily.dev._ ## Similar posts on daily.dev - [How I shipped a Temporal-powered product in three weeks, one year out of college](https://daily.dev/posts/how-i-shipped-a-temporal-powered-product-in-three-weeks-one-year-out-of-college-qeoyjiylf) · Temporal · 0 upvotes · 0 comments - [Claude Code cache confusion as Anthropic tweaks defaults, but quotas still drain](https://daily.dev/posts/claude-code-cache-confusion-as-anthropic-tweaks-defaults-but-quotas-still-drain-uvn7zes3k) · DEVCLASS · 5 upvotes · 1 comments - [The Cost of Proof Has Collapsed](https://daily.dev/posts/the-cost-of-proof-has-collapsed-nflsmuiln) · Medium · 0 upvotes · 0 comments --- Tags: [#recommendation-systems](https://daily.dev/tags/recommendation-systems) [View this post on daily.dev](https://daily.dev/posts/airbnb-cuts-recommendation-staleness-from-2-days-to-seconds-with-chronon-xwnl5ftip) ```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":"Airbnb Cuts Recommendation Staleness from 2 Days to Seconds with Chronon","url":"https://daily.dev/posts/airbnb-cuts-recommendation-staleness-from-2-days-to-seconds-with-chronon-xwnl5ftip","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/airbnb-cuts-recommendation-staleness-from-2-days-to-seconds-with-chronon-xwnl5ftip"},"datePublished":"2026-10-28T14:05:47.727Z","dateModified":"2026-10-28T14:57:21.585Z","description":"Airbnb extended its open source Chronon feature platform with Push Mode and Near Real Time Model Transform to refresh recommendation embeddings within seconds...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/215537ba54a899534f5544933a39f181?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/215537ba54a899534f5544933a39f181?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"InfoQ","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":"InfoQ","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/afc3bced3e1e4b188dd9127017a60e0c","url":"https://daily.dev/sources/infoq"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/airbnb-cuts-recommendation-staleness-from-2-days-to-seconds-with-chronon-xwnl5ftip","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"recommendation-systems","timeRequired":"PT3M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"InfoQ","item":"https://daily.dev/sources/infoq"},{"@type":"ListItem","position":3,"name":"Airbnb Cuts Recommendation Staleness from 2 Days to Seconds with Chronon"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/airbnb-cuts-recommendation-staleness-from-2-days-to-seconds-with-chronon-xwnl5ftip#faq","mainEntity":[{"@type":"Question","name":"How did Airbnb reduce staleness in its recommendation embeddings using Chronon?","acceptedAnswer":{"@type":"Answer","text":"Airbnb added Push Mode and Near Real-Time Model Transform to its Chronon feature platform, letting a streaming job publish a notification when a feature value changes and run the already-deployed Transformer sequence encoder inside that streaming pipeline. This cut embedding staleness from nearly two days to under a minute, typically 10 to 30 seconds. Developers designing low-latency feature pipelines can follow architecture breakdowns like this one on daily.dev."}},{"@type":"Question","name":"What business impact did Airbnb see from switching to near real-time embeddings in Chronon?","acceptedAnswer":{"@type":"Answer","text":"Airbnb measured a 1.67% improvement in offline NDCG and an approximately 0.33% increase in uncancelled bookings during online A/B testing after deploying near real-time embedding updates through Chronon's Push Mode and NRT Model Transform. Teams justifying real-time ML infrastructure investments can track results like these on daily.dev."}},{"@type":"Question","name":"What is NRT Model Transform in Airbnb's Chronon feature platform?","acceptedAnswer":{"@type":"Answer","text":"NRT Model Transform is a Chronon capability that lets an already deployed model run directly inside a streaming pipeline, so the same Transformer sequence encoder used for daily batch processing can also generate embeddings in near real time, avoiding a separate inference implementation for the streaming workflow. Engineers evaluating streaming inference architectures can follow platform updates like this on daily.dev."}}]} ```