20.09.2024 aktualisiert
100 % verfügbar
Software Engineer
Hamburg, Deutschland
Weltweit
Skills
JavascriptDeploymentFrontendBackendNodeJSFrontend-DeveloperMEAN stackMobile WebseitenWebsiteGolangRustTypescriptMongoDBawsAmazon Web ServiceJavaLambdaCDKDynamoDBS3 Object Storageapi gatewayAWS VPC
Plenty of experience in Software Engineering, Distributed Systems, Cloud Infrastructure and Operational Excellence.
Just done 4 years at Amazon S3 as a Software Engineer operating the largest storage system in the world.
Experience in mentoring and setting up more junior colleagues for success.
Just done 4 years at Amazon S3 as a Software Engineer operating the largest storage system in the world.
Experience in mentoring and setting up more junior colleagues for success.
Sprachen
DeutschMutterspracheEnglischverhandlungssicherPortugiesischverhandlungssicherSpanischgut
Projekthistorie
At AWS I have been working on the worlds largest Storage System - Amazon S3. In particular my team and I owned the "Intelligent Tiering"-Storage Class which automatically tiers customer blobs based on their access patterns to achieve best storage-to-access cost ratios. Doing so our system managed hundreds of billions of customer blobs. This involved big data handling on spark clusters and a tiering services. The most important aspect of all in S3 is the operational excellence. Our systems achieve 5 9's of availability!
I was also part of a task force (that later evolved into a full fledged team) that was was dedicated to reducing the IO demand in the entire S3 system by up to 10% enabling hundreds of million of dollars CapEx savings. This was achieved by leveraging machine learning to intelligently influence blob placement. This again required enormous large scale engineering and had only a performance budget of <5ms.
I was also part of a task force (that later evolved into a full fledged team) that was was dedicated to reducing the IO demand in the entire S3 system by up to 10% enabling hundreds of million of dollars CapEx savings. This was achieved by leveraging machine learning to intelligently influence blob placement. This again required enormous large scale engineering and had only a performance budget of <5ms.