赞助方：Co​​alesce 化乱为治：如何在数据与分析领域取得成功.pdf

1、Bringing Order to Chaos:How to Succeed in a Data&Analytics World CoalesceBooth 319Speaker 3Michael TantrumSales Engineering Director at CoalesceThe data persons raison dtre“Our jobs exist to get data in the hands of business users so they can manage their business.”Data must be:TimelyEasily understo

2、od and fit for purposeAccurate and of good quality4Why is this so hard?Data projects historically have a tremendously high failure rateCommon reasons:Change management failuresWeak/constantly changing requirementsCumbersome processesData challenges5Change management“Axiom of Data Projects:Consumers

3、of data are unable to articulate their requirements clearly-they dont know what they dont know”“Corollary:They will know it when they see it”Data Engineering is the only branch of engineering where the full specifications of the deliverable are unknown at the start of the projectTherefore:Standard e

4、ngineering processes will not work for Data Engineering6Typical data engineering development cycle7Business NeedRequirements DefinitionBuildAcceptanceRemediationDeployTypical data engineering development cycle8Business NeedRequirements DefinitionBuildAcceptanceRemediationDeployOur jobTypical data en

5、gineering development cycle9Business NeedRequirements DefinitionBuildAcceptanceRemediationOur jobSomeone elses jobDeployTypical data engineering development cycle10Business NeedRequirements DefinitionBuildAcceptanceRemediationOur jobSomeone elses jobFundamental DisconnectWe get stuck here DeployThe

6、resultWe burn out and take shortcuts just to get the project doneWe hate our users,because they keep making changesThey hate us,because we appear to be the bottleneck,take too long,and are apparently inflexibleThe project never gets completed to everyones satisfactionCost and tim