Comcast Effectv 如何通过 Databricks 和 Monte Carlo 推动数据可观测性.pdf

《Comcast Effectv 如何通过 Databricks 和 Monte Carlo 推动数据可观测性.pdf》由会员分享，可在线阅读，更多相关《Comcast Effectv 如何通过 Databricks 和 Monte Carlo 推动数据可观测性.pdf（40页珍藏版）》请在三个皮匠报告上搜索。

1、How Comcasts Effectv Drives Data Observability with Databricks and Monte CarloDatabricks20232IntroductionsSCOTT LERNER Customer Success Manager,Monte CarloROBINSON CREIGHTONSr.Manager DataOps,Effectv1_DAIS_Title_SlideWhat is Monte Carlo?What is Data Observability?Good pipelines,bad data Is the data

2、up-to-date?Is the data complete?Are fields within expected ranges?Is the null rate higher or lower than it should be?Has the schema changed?and many moreDATA OBSERVABILITY PILLARSFreshnessVolumeSchemaQualityLineageWhat is Data Observability?Good pipelines,bad data PreventAuto-generated and on-demand

3、 insights Schema change notificationsAutomated circuit breakersDetectML-powered anomaly detectionRule-based detectionTargeted alerts to impacted owners&downstream usersResolveAutomated field-level lineage Impact radius assessmentCode,data,and operational diagnostics DATA OBSERVABILITY PILLARSFreshne

4、ss|Volume|Quality|Schema|LineageEffectv:Who We Are&What We DoEffectv is an audience deliveryaudience deliverycompany that combines the best of digitalbest of digital with the power of TVpower of TVData at EffectvAd supported by Effectv|Not ad supported by Effectv|Not ad supported at allat all5 billi

5、on5 billion streams monthly across our platforms60+Markets60+MarketsWith geographic targeting 170170Networks 11,00011,000TV programs2.5 Billion2.5 Billion2Average viewing hours per month30 Million30 MillionComcast subscriber households96 Million96 Million1Estimated audience of adults reached20+20+Da

6、ta partnerships Source:1.Estimate based on U.S.census of broadband subscriber households in Comcast-represented U.S.counties 2.Comcast Aggregated Viewership Data.2022 Monthly Averages&Full Footprint.Data Domains to Business ValueSegment data using Monte Carlo domains CUSTOMERCUSTOMEROrganizations th