{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T15:57:18Z","timestamp":1780588638278,"version":"3.54.1"},"reference-count":52,"publisher":"Association for Computing Machinery (ACM)","issue":"3","content-domain":{"domain":["dl.acm.org"],"crossmark-restriction":true},"short-container-title":["Proc. VLDB Endow."],"published-print":{"date-parts":[[2023,11]]},"abstract":"<jats:p>\n            Feature stores (also sometimes referred to as embedding stores) are becoming ubiquitous in model serving systems: downstream applications query these stores for auxiliary inputs at inference-time. Stored features are derived by\n            <jats:italic>featurizing<\/jats:italic>\n            rapidly changing base data sources. Featurization can be costly prohibitively expensive to trigger on every data update, particularly for features that are vector embeddings computed by a model. Yet, existing systems naively apply a one-size-fits-all policy as to when\/how to update these features, and do not consider query access patterns or impacts on prediction accuracy. This paper introduces RALF, which orchestrates feature updates by leveraging\n            <jats:italic>downstream error feedback<\/jats:italic>\n            to minimize\n            <jats:italic>feature store regret<\/jats:italic>\n            , a metric for how much featurization degrades downstream accuracy. We evaluate with representative feature store workloads, anomaly detection and recommendation, using real-world datasets. We run system experiments with a 275,077 key anomaly detection workload on 800 cores to show up to a 32.7% reduction in prediction error or up to 1.6X compute cost reduction with accuracy-aware scheduling.\n          <\/jats:p>","DOI":"10.14778\/3632093.3632116","type":"journal-article","created":{"date-parts":[[2024,1,20]],"date-time":"2024-01-20T11:26:31Z","timestamp":1705749991000},"page":"563-576","update-policy":"https:\/\/doi.org\/10.1145\/crossmark-policy","source":"Crossref","is-referenced-by-count":6,"title":["RALF: Accuracy-Aware Scheduling for Feature Store Maintenance"],"prefix":"10.14778","volume":"17","author":[{"given":"Sarah","family":"Wooders","sequence":"first","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Xiangxi","family":"Mo","sequence":"additional","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Amit","family":"Narang","sequence":"additional","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kevin","family":"Lin","sequence":"additional","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ion","family":"Stoica","sequence":"additional","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joseph M.","family":"Hellerstein","sequence":"additional","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Natacha","family":"Crooks","sequence":"additional","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Joseph E.","family":"Gonzalez","sequence":"additional","affiliation":[{"name":"UC Berkeley"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"320","published-online":{"date-parts":[[2024,1,20]]},"reference":[{"key":"e_1_2_1_1_1","volume-title":"Amazon Web Services. Retrieved","author":"Amazon Web","year":"2023","unstructured":"Amazon Web Services [n.d.]. Amazon SageMaker Feature Store. Amazon Web Services. Retrieved November 11, 2023 from https:\/\/aws.amazon.com\/sagemaker\/feature-store\/"},{"key":"e_1_2_1_2_1","volume-title":"Hopsworks. Retrieved","author":"Hopsworks","year":"2023","unstructured":"Hopsworks [n.d.]. Feature Stores Org. Hopsworks. Retrieved November 11, 2023 from https:\/\/www.featurestore.org\/"},{"key":"e_1_2_1_3_1","volume-title":"Feature Tools. Retrieved","year":"2023","unstructured":"[n.d.]. Feature Tools. Retrieved November 11, 2023 from https:\/\/featuretools.alteryx.com\/en\/stable\/getting_started\/handling_time.html"},{"key":"e_1_2_1_4_1","unstructured":"Hopsworks [n.d.]. Hopsworks. Hopsworks. Retrieved November 11 2023 from https:\/\/www.hopsworks.ai\/"},{"key":"e_1_2_1_5_1","volume-title":"grouplens. Retrieved","author":"Dataset MovieLens","year":"2023","unstructured":"grouplens [n.d.]. MovieLens 1M Dataset. grouplens. Retrieved November 11, 2023 from https:\/\/grouplens.org\/datasets\/movielens\/1m\/"},{"key":"e_1_2_1_6_1","volume-title":"Rate Limits. OpenAI. Retrieved","author":"AI","year":"2023","unstructured":"OpenAI [n.d.]. Rate Limits. OpenAI. Retrieved November 11, 2023 from https:\/\/platform.openai.com\/docs\/guides\/rate-limits?context=tier-free"},{"key":"e_1_2_1_7_1","unstructured":"[n.d.]. Redis. Retrieved November 11 2023 from https:\/\/redis.io\/"},{"key":"e_1_2_1_8_1","volume-title":"affordable pricing. Cohere. Retrieved","author":"Scalable Cohere","year":"2023","unstructured":"Cohere [n.d.]. Scalable, affordable pricing. Cohere. Retrieved November 11, 2023 from https:\/\/cohere.com\/pricing"},{"key":"e_1_2_1_9_1","unstructured":"Tecton [n.d.]. Tecton. Tecton. Retrieved November 11 2023 from https:\/\/www.tecton.ai\/"},{"key":"e_1_2_1_10_1","unstructured":"[n.d.]. Timely Dataflow. Retrieved November 11 2023 from https:\/\/timelydataflow.github.iASo\/timely-dataflow\/"},{"key":"e_1_2_1_11_1","volume-title":"Tensorflow: A system for large-scale machine learning. In 12th {USENIX} symposium on operating systems design and implementation ({OSDI} 16). 265--283.","author":"Abadi Mart\u00edn","year":"2016","unstructured":"Mart\u00edn Abadi, Paul Barham, Jianmin Chen, Zhifeng Chen, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Geoffrey Irving, Michael Isard, et al. 2016. Tensorflow: A system for large-scale machine learning. In 12th {USENIX} symposium on operating systems design and implementation ({OSDI} 16). 265--283."},{"key":"e_1_2_1_12_1","doi-asserted-by":"publisher","DOI":"10.1145\/2465351.2465355"},{"key":"e_1_2_1_13_1","doi-asserted-by":"publisher","DOI":"10.14778\/2735508.2735509"},{"key":"e_1_2_1_14_1","doi-asserted-by":"publisher","DOI":"10.14778\/2212351.2212359"},{"key":"e_1_2_1_15_1","volume-title":"Ekya: Continuous Learning of Video Analytics Models on Edge Compute Servers. arXiv preprint arXiv:2012.10557","author":"Bhardwaj Romil","year":"2020","unstructured":"Romil Bhardwaj, Zhengxu Xia, Ganesh Ananthanarayanan, Junchen Jiang, Nikolaos Karianakis, Yuanchao Shu, Kevin Hsieh, Victor Bahl, and Ion Stoica. 2020. Ekya: Continuous Learning of Video Analytics Models on Edge Compute Servers. arXiv preprint arXiv:2012.10557 (2020)."},{"key":"e_1_2_1_16_1","doi-asserted-by":"publisher","DOI":"10.1145\/3035918.3056097"},{"key":"e_1_2_1_17_1","doi-asserted-by":"publisher","DOI":"10.1561\/1900000020"},{"key":"e_1_2_1_18_1","doi-asserted-by":"publisher","DOI":"10.14778\/1454159.1454167"},{"key":"e_1_2_1_19_1","doi-asserted-by":"publisher","DOI":"10.1145\/3132747.3132772"},{"key":"e_1_2_1_20_1","doi-asserted-by":"publisher","DOI":"10.1145\/3419111.3421285"},{"key":"e_1_2_1_21_1","volume-title":"Clipper: A low-latency online prediction serving system. In 14th {USENIX} Symposium on Networked Systems Design and Implementation ({NSDI} 17). 613--627.","author":"Crankshaw Daniel","year":"2017","unstructured":"Daniel Crankshaw, Xin Wang, Guilio Zhou, Michael J Franklin, Joseph E Gonzalez, and Ion Stoica. 2017. Clipper: A low-latency online prediction serving system. In 14th {USENIX} Symposium on Networked Systems Design and Implementation ({NSDI} 17). 613--627."},{"key":"e_1_2_1_22_1","doi-asserted-by":"publisher","DOI":"10.1145\/2882903.2882951"},{"key":"e_1_2_1_23_1","volume-title":"Proceedings of the 2014 USENIX Conference on USENIX Annual Technical Conference (Philadelphia, PA) (USENIX ATC'14). USENIX Association, USA, 37--48","author":"Cui Henggang","unstructured":"Henggang Cui, James Cipar, Qirong Ho, Jin Kyu Kim, Seunghak Lee, Abhimanu Kumar, Jinliang Wei, Wei Dai, Gregory R. Ganger, Phillip B. Gibbons, Garth A. Gibson, and Eric P. Xing. 2014. Exploiting Bounded Staleness to Speed up Big Data Analytics. In Proceedings of the 2014 USENIX Conference on USENIX Annual Technical Conference (Philadelphia, PA) (USENIX ATC'14). USENIX Association, USA, 37--48."},{"key":"e_1_2_1_24_1","volume-title":"Martin Ek, Eddie Kohler, M Frans Kaashoek, and Robert Morris.","author":"Gjengset Jon","year":"2018","unstructured":"Jon Gjengset, Malte Schwarzkopf, Jonathan Behrens, Lara Timb\u00f3 Ara\u00fajo, Martin Ek, Eddie Kohler, M Frans Kaashoek, and Robert Morris. 2018. Noria: dynamic, partially-stateful data-flow for high-performance web applications. In 13th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 18). 213--231."},{"key":"e_1_2_1_25_1","volume-title":"International conference on machine learning. PMLR, 3929--3938","author":"Guu Kelvin","year":"2020","unstructured":"Kelvin Guu, Kenton Lee, Zora Tung, Panupong Pasupat, and Mingwei Chang. 2020. Retrieval augmented language model pre-training. In International conference on machine learning. PMLR, 3929--3938."},{"key":"e_1_2_1_26_1","doi-asserted-by":"publisher","DOI":"10.1145\/2987550.2987559"},{"key":"e_1_2_1_27_1","doi-asserted-by":"publisher","DOI":"10.1145\/3230543.3230574"},{"key":"e_1_2_1_28_1","volume-title":"Proc. 2nd SysML Conf.","author":"Kakantousis Theofilos","year":"2019","unstructured":"Theofilos Kakantousis, Antonios Kouzoupis, Fabio Buso, Gautier Berthou, Jim Dowling, and Seif Haridi. 2019. Horizontally scalable ml pipelines with a feature store. In Proc. 2nd SysML Conf., Palo Alto, USA."},{"key":"e_1_2_1_29_1","volume-title":"Noscope: optimizing neural network queries over video at scale. arXiv preprint arXiv:1703.02529","author":"Kang Daniel","year":"2017","unstructured":"Daniel Kang, John Emmons, Firas Abuzaid, Peter Bailis, and Matei Zaharia. 2017. Noscope: optimizing neural network queries over video at scale. arXiv preprint arXiv:1703.02529 (2017)."},{"key":"e_1_2_1_30_1","volume-title":"Approximate selection with guarantees using proxies. arXiv preprint arXiv:2004.00827","author":"Kang Daniel","year":"2020","unstructured":"Daniel Kang, Edward Gan, Peter Bailis, Tatsunori Hashimoto, and Matei Zaharia. 2020. Approximate selection with guarantees using proxies. arXiv preprint arXiv:2004.00827 (2020)."},{"key":"e_1_2_1_31_1","doi-asserted-by":"publisher","DOI":"10.1109\/MC.2009.263"},{"key":"e_1_2_1_32_1","volume-title":"Willump: A statistically-aware end-to-end optimizer for machine learning inference. arXiv preprint arXiv.1906.01974","author":"Kraft Peter","year":"2019","unstructured":"Peter Kraft, Daniel Kang, Deepak Narayanan, Shoumik Palkar, Peter Bailis, and Matei Zaharia. 2019. Willump: A statistically-aware end-to-end optimizer for machine learning inference. arXiv preprint arXiv.1906.01974 (2019)."},{"key":"e_1_2_1_33_1","doi-asserted-by":"publisher","DOI":"10.1145\/2465351.2465363"},{"key":"e_1_2_1_34_1","unstructured":"N Laptev and S Amizadeh. 2015. Yahoo anomaly detection dataset s5. http:\/\/webscope.sandbox.yahoo.com\/catalog.php (2015)."},{"key":"e_1_2_1_35_1","volume-title":"Latent retrieval for weakly supervised open domain question answering. arXiv preprint arXiv.1906.00300","author":"Lee Kenton","year":"2019","unstructured":"Kenton Lee, Ming-Wei Chang, and Kristina Toutanova. 2019. Latent retrieval for weakly supervised open domain question answering. arXiv preprint arXiv.1906.00300 (2019)."},{"key":"e_1_2_1_36_1","first-page":"9459","article-title":"Retrieval-augmented generation for knowledge-intensive nlp tasks","volume":"33","author":"Lewis Patrick","year":"2020","unstructured":"Patrick Lewis, Ethan Perez, Aleksandra Piktus, Fabio Petroni, Vladimir Karpukhin, Naman Goyal, Heinrich K\u00fcttler, Mike Lewis, Wen-tau Yih, Tim Rockt\u00e4schel, et al. 2020. Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems 33 (2020), 9459--9474.","journal-title":"Advances in Neural Information Processing Systems"},{"key":"e_1_2_1_37_1","doi-asserted-by":"publisher","DOI":"10.5555\/2387880.2387906"},{"key":"e_1_2_1_38_1","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58536-5_28"},{"key":"e_1_2_1_39_1","doi-asserted-by":"publisher","DOI":"10.1145\/2043556.2043593"},{"key":"e_1_2_1_40_1","doi-asserted-by":"publisher","DOI":"10.14778\/3401960.3401974"},{"key":"e_1_2_1_41_1","volume-title":"OnlineSTL: Scaling Time Series Decomposition by 100x. arXiv e-prints","author":"Mishra Abhinav","year":"2021","unstructured":"Abhinav Mishra, Ram Sriharsha, and Sichen Zhong. 2021. OnlineSTL: Scaling Time Series Decomposition by 100x. arXiv e-prints (2021), arXiv-2107."},{"key":"e_1_2_1_42_1","volume-title":"Ray: A distributed framework for emerging {AI} applications. In 13th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 18). 561--577.","author":"Moritz Philipp","year":"2018","unstructured":"Philipp Moritz, Robert Nishihara, Stephanie Wang, Alexey Tumanov, Richard Liaw, Eric Liang, Melih Elibol, Zongheng Yang, William Paul, Michael I Jordan, et al. 2018. Ray: A distributed framework for emerging {AI} applications. In 13th {USENIX} Symposium on Operating Systems Design and Implementation ({OSDI} 18). 561--577."},{"key":"e_1_2_1_43_1","unstructured":"Maxim Naumov Dheevatsa Mudigere Hao-Jun Michael Shi Jianyu Huang Narayanan Sundaraman Jongsoo Park Xiaodong Wang Udit Gupta Carole-Jean Wu Alisson G. Azzolini Dmytro Dzhulgakov Andrey Mallevich Ilia Cherniavskii Yinghai Lu Raghuraman Krishnamoorthi Ansha Yu Volodymyr Kondratenko Stephanie Pereira Xianjie Chen Wenlin Chen Vijay Rao Bill Jia Liang Xiong and Misha Smelyanskiy. 2019. Deep Learning Recommendation Model for Personalization and Recommendation Systems. CoRR abs\/1906.00091 (2019). https:\/\/arxiv.org\/abs\/1906.00091"},{"key":"e_1_2_1_44_1","volume-title":"MemGPT: Towards LLMs as Operating Systems. arXiv preprint arXiv.2310.08560","author":"Packer Charles","year":"2023","unstructured":"Charles Packer, Vivian Fang, Shishir G Patil, Kevin Lin, Sarah Wooders, and Joseph E Gonzalez. 2023. MemGPT: Towards LLMs as Operating Systems. arXiv preprint arXiv.2310.08560 (2023)."},{"key":"e_1_2_1_45_1","doi-asserted-by":"publisher","DOI":"10.25080\/Majora-92bf1922-011"},{"key":"e_1_2_1_46_1","volume-title":"Proceedings of the 2012 ACM SIGMOD International Conference on Management of Data. 729--730 year=2012","author":"Sivasubramanian Swaminathan","unstructured":"Swaminathan Sivasubramanian. [n.d.]. Amazon dynamoDB: a seamlessly scalable non-relational database service. In Proceedings of the 2012 ACM SIGMOD International Conference on Management of Data. 729--730 year=2012."},{"key":"e_1_2_1_47_1","doi-asserted-by":"publisher","DOI":"10.1145\/2517349.2522731"},{"key":"e_1_2_1_48_1","volume-title":"2021 International Conference on Data Mining Workshops (ICDMW). IEEE, 414--420","author":"Wang Zhaohui","year":"2021","unstructured":"Zhaohui Wang, Xiao Lin, Abhinav Mishra, and Ram Sriharsha. 2021. Online Changepoint Detection on a Budget. In 2021 International Conference on Data Mining Workshops (ICDMW). IEEE, 414--420."},{"key":"e_1_2_1_49_1","doi-asserted-by":"publisher","DOI":"10.1145\/137097.137880"},{"key":"e_1_2_1_50_1","doi-asserted-by":"publisher","unstructured":"Kun-Lung Wu P.S. Yu and C. Pu. 1992. Divergence control for epsilon-serializability. In [1992] Eighth International Conference on Data Engineering. 506--515. 10.1109\/ICDE.1992.213158","DOI":"10.1109\/ICDE.1992.213158"},{"key":"e_1_2_1_51_1","volume-title":"Proceedings of the 4th Conference on Symposium on Operating System Design & Implementation -","volume":"4","author":"Yu Haifeng","year":"2000","unstructured":"Haifeng Yu and Amin Vahdat. 2000. Design and Evaluation of a Continuous Consistency Model for Replicated Services. In Proceedings of the 4th Conference on Symposium on Operating System Design & Implementation - Volume 4 (San Diego, California) (OSDI'00). USENIX Association, USA, Article 21."},{"key":"e_1_2_1_52_1","unstructured":"Jingren Zhou Per-ke Larson and Jonathan Goldstein. 2005. Partially materialized views. In submitted to this conference."}],"container-title":["Proceedings of the VLDB Endowment"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/dl.acm.org\/doi\/pdf\/10.14778\/3632093.3632116","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,1,20]],"date-time":"2024-01-20T11:28:58Z","timestamp":1705750138000},"score":1,"resource":{"primary":{"URL":"https:\/\/dl.acm.org\/doi\/10.14778\/3632093.3632116"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2023,11]]},"references-count":52,"journal-issue":{"issue":"3","published-print":{"date-parts":[[2023,11]]}},"alternative-id":["10.14778\/3632093.3632116"],"URL":"https:\/\/doi.org\/10.14778\/3632093.3632116","relation":{},"ISSN":["2150-8097"],"issn-type":[{"value":"2150-8097","type":"print"}],"subject":[],"published":{"date-parts":[[2023,11]]},"assertion":[{"value":"2024-01-20","order":2,"name":"published","label":"Published","group":{"name":"publication_history","label":"Publication History"}}]}}