This Is My Architecture (TMA) is a new video series that highlights innovative architectural solutions on the AWS Cloud by AWS customers and partners. These 2-5-minute video segments are designed for a technical audience and showcase the most interesting and technically creative elements of each architecture.

Subscribe to the YouTube playlist. We'll be adding new videos on a regular basis, so check back frequently!

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-infor
3:59

David Lowry, Sr. Cloud Manager at Infor, explains how they built Windsor, a serverless application testing solution. Infor uses Windsor to set up, tear down, and run test cases on over 50 multi-tenant enterprise applications that are deployed on AWS. It is built with a combination of Amazon SWF, AWS Lambda, Amazon DynamoDB, and Amazon CloudWatch Events. This serverless approach to application testing costs Infor about $1 a month, or less than a cup of coffee!

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-filemaker
5:11

In this episode of This Is My Architecture, James Qiu, Director of Engineering at FileMaker, describes how they built a solution for SMB customers that automatically scales FileMaker to handle surge events. The solution leverages the AWS Marketplace, where FileMaker lists its software, in addition to AWS CloudFormation to deploy and resize FileMaker environments on AWS as needed. This flexibility allows customers to easily scale the infrastructure powering their FileMaker solution on AWS up or down to handle varying levels of load.

View the architecture diagram. View the FileMaker on-premise vs. cloud/AMI architecture.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-liveramp
4:52

In this episode of This Is My Architecture, Michel from LiveRamp walks us through their end-to-end streaming data processing platform that handles 350,000 requests per second and 15-17TB of logs per day. You'll learn how they leverage Amazon EC2, Auto Scaling, Amazon Kinesis Firehose, Amazon S3, Amazon EMR, and Amazon Redshift to power a fast, scalable, and reliable marketing data analytics platform that meets strict SLAs.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-expedia
7:13

Kuldeep Chowhan, Principal Engineer at Expedia, shows us how Expedia built a scalable CI/CD platform on AWS using Github, Jenkins, Amazon EC2 Container Service (ECS), and Lambda. With over 4,000 applications deployed through this platform, their developers can deploy a wide variety of apps faster and more frequently. The platform gathers and stores rollout metrics and data, and with Blue-Green deployment support, it supports continuous improvement in their production environments.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-fanduel
5:02

In this episode of This is My Architecture, Steve Hunt, Director of Infrastructure and Ops at Fanduel, shows how they run CloudFormation at scale. With over 400 stacks running at any given time, you'll see how they architected their tooling to manage this infrastructure. You'll also see how they scale at the data layer, using services like Amazon Aurora and Amazon ElastiCache, along with techniques like precomputed cache to deliver optimal data services to their stacks.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-spirelabs
9:08

In this episode of This Is My Architecture, Rob Scott (@robertjscott), VP of Software Architecture at Spire Labs (@spire), explains how they built a fault-tolerant solution across three AWS Availability Zones with Amazon RDS, Amazon ElastiCache, and Kubernetes on Amazon EC2 with Auto Scaling. Their design allows them to tolerate the temporary loss of both Kubernetes pods/containers and EC2 instances, keeping them up and running smoothly for customers.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-clearscale
3:40

In this episode of This Is My Architecture, Jimmy Chui from ClearScale explains how they built a platform that allows customers to order "designer genes". Their solution uses serverless products such as Amazon API Gateway, Amazon Lambda, Amazon DynamoDB, and Amazon SQS, in addition to Amazon EC2 Container Service (ECS) clusters to run and manage their Docker containers. The ClearScale platform is built to be flexible and scalable: it handles 1000 requests per gene sequence, with spiky workloads that peak 25-30% of the day.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-idt
2:46

In this episode of This Is My Architecture, Nathaniel Ritholtz, Software Developer at IDT, describes how they built a secure configuration management modification system for their containerized applications on AWS. IDT's solution uses HashiCorp's Consul along with Amazon Simple Storage Service (S3), AWS Key Management Service (KMS), and Amazon EC2 Container Service (ECS).

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-deloitte
6:16

In this episode of This is My Architecture, Fran Cardell, Cloud Hosting Lead at Deloitte, explains how they simplified running SAP Hana on AWS. Starting with a Quick Start reference architecture, they build a cost-optimized, well-performing platform with robust security controls. You'll also hear about the storage and instance types that go into building small, medium, and large versions of the platform.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-dynatrace
5:38

Andreas Lehofer, Chief Product Officer at Dynatrace, explains how they built their Application Performance Management Solution by securely using four different types of storage on AWS. You'll learn how they use Lambda to securely ingest logs and performance data at high speed. You'll also learn about davis, a virtual assistant for digital performance management, which enables VoiceOps through natural language queries to Alexa and artificial intelligence running on AWS. Invite your infrastructure assistant to your next stand-up meeting!

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-irobot
5:50

Ben Kehoe, Cloud Robotics Research Scientist at iRobot, explains how they built a serverless solution to power microservices that scale to control millions of robots. You’ll learn how they automate and optimize AWS Lambda functions and Amazon API Gateway deployments with AWS CloudFormation and Swagger, plus how they inject information during the deployment process to decouple architecture details from the code.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-airbnb
3:39

In this episode of This is My Architecture, Xinyao Hu, Engineering Manager at Airbnb, discusses the Amazon RDS Replay Machine they built. They built an infrastructure that allows them to replay actual transactions sent to almost 100 different databases. Tune in and see how Airbnb uses this system for disaster recovery, load testing, and advanced analytics in Hive at massive scale.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-coinbase2
1:55

In this episode of This is My Architecture, Rob Witoff, Director of Infrastructure, dives into AWS IAM and AWS CloudTrail policy management at Coinbase. You'll see how they use the IAM Access Advisor to descope permissions in Development before pushing to Production. You'll also learn how they wrap their open-source tool GeoEngineer around IAM and CloudTrail to maintain their high-security environment without inhibiting their Operations teams.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-human-longevity
4:20

Bryan Coon, Head of Enterprise Services at Human Longevity Inc., explains how they process up to 12TB per day of raw data in Amazon S3 with custom analytics tools running in Docker containers. HLI has automated the deployment of reusable pipelines of containerized tasks with Amazon SWF, the Ruby Flow framework, AWS OpsWorks, Amazon SQS, and AWS Lambda. They also keep costs low by leveraging Amazon EC2 Spot Fleet to power their Amazon ECS clusters.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-station-x
5:17

In this episode of This Is My Architecture, Anish from Station X explains how they built a platform to analyze, visualize, and manage genomic information at scale on AWS. You’ll learn how they use Qubole to manage clusters running Presto and Hive that power their interactive, near real-time query interface. You’ll also learn how they optimize query performance by storing their data in ORC stripes and sorting by genomic coordinate.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-capital-one
5:55

In this episode of This is My Architecture, Kapil Thangavelu, Director of Engineering at Capital One, explains their open-sourced rules engine for fleet management: Cloud Custodian. You'll see how Capital One uses this tool and a simple YAML-based DSL to eliminate the need to manage hundreds or thousands of scripts and policies, and offer "real-time" compliance and cost management at scale.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-vtex
5:00

In this episode of This is My Architecture, Geraldo Thomaz, Co-CEO at VTEX, an AWS advanced partner, explains how they built a high security SaaS environment using IAM and MFA federated access. You’ll see how their continuous delivery model uses SAML integration and the AWS Secure Token Service to provide temporary keys to maintain least privilege.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-heroku
3:33

Greg Nokes from Heroku explains how one of their customers built an Internet of Things (IoT) solution that uses Heroku’s managed Kafka service, Amazon S3 for data storage, and Amazon Redshift for analytics.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-new-relic
4:59

In this episode of This Is My Architecture, Lisa Moreno from New Relic explains how they take a modular approach to automated AWS infrastructure deployment using HashiCorp's Terraform.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-corpinfo
6:00

In this episode of This Is My Architecture, Tim Fox from CorpInfo, an AWS Premier Consulting Partner, explains how they built a flexible and scalable IoT solution to power a customer's video streaming application. You'll see how they use a mix of AWS services and data stores, including ELB, EC2, S3, RDS, DynamoDB, and Redshift, plus the concept of "pods" to optimize for a variety of different requirements.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-civitas-learning
2:59

In this episode of This Is My Architecture, Chad Walters, VP of Engineering at Civitas Learning, explains how they spin-up transient, parallel Amazon Redshift clusters to securely separate PII data. He also explains how they use an Amazon EC2-based Scala application called "Foreman" to dynamically right-size Redshift clusters based on the datasets' metadata in addition to past query performance.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-ibm-spectrum-protect
4:04

In this episode of This is My Architecture, Jay Calder, Lead Cloud Developer for IBM Spectrum Protect, dives in to how they built an incremental forever architecture on top on Amazon S3.

To learn more, read this case study.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-financial-times
3:40

Greg Cope, Head of Platform Architecture and Security, explains how the team at Financial Times used Cloudhealth, Lambda, and shared dashboards to encourage the adoption of Amazon Linux and gamify cost optimization. It's the Thames meets the Silicon Valley in this episode of This is My Architecture.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-ironsource
6:16

How do you get near real-time insight when your data sources are all over the planet? Shimon Tolts, AWS Community Hero and CTO at ironSource, shares how they built ironSource Atom using Amazon Kinesis, Amazon Redshift, and geo-latency based routing to collect and process over 160 billion events per month. With this architecture, their customers can see enriched data in their data warehouse in about a minute.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-eseye
2:51

You want secure communication to a device in the field, but how do you make that first connection without touching the device? Ian Marsden, CTO of Eseye (an AWS IoT Competency partner) shares how they built AnyNet Secure for AWS IoT to solve this problem. This global cellular connectivity product enables AWS customers to automatically activate, provision, authenticate, and certify IoT devices, and then ingest data to their AWS Cloud in over 153 countries.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-coinbase
5:28

Coinbase is the easiest place to buy and sell cryptocurrencies, so security is built into everything they do. In this episode of This is My Architecture, Rob Witoff, Director of Infrastructure at Coinbase, explains how they empower their developers to make changes by requiring consensus to actually make those changes. You'll learn how they use their open-source tool GeoEngineer on top of Terraform to make production changes quickly and securely.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-netflix
4:14

What happens when you need to move 89 million viewers to a different AWS Region? Netflix's infrastructure, built on AWS, makes it possible to be extremely resilient, even when you're running services in many AWS Regions simultaneously. Coburn Watson, Director of Performance and Reliability Engineering at Netflix, takes us through a DNS architecture—built on Amazon Route 53 and augmented with Netflix's Zuul—that allows them to evacuate an entire region in under 40 minutes.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-sumologic
6:03

Stefan Zier, Chief Architect at Sumo Logic, explains how their solution leverages multiple AWS services to ingest and process massive amounts of logs for their customers. You'll learn about how they use Amazon Simple Storage Service (Amazon S3) to store logs and state, provide customer-specific encryption with keys that rotate daily, and leverage lifecycle policy and the infrequent access data class to save money. Stefan also reviews how Sumo Logic uses clusters of EC2 instances to run Kafka and other software in parallel workflows with low end-to-end latency, and Amazon DynamoDB's atomic counter feature to meter usage so that they can accurately charge their customers.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-checkpoint-architecture
4:17

Greg Pepper, Security Architect at Check Point Software Technologies, explains how they built their AWS solution to automatically scale both vertically and horizontally. You'll learn about how they leverage Auto Scaling and the "ELB sandwich" design pattern in an Amazon Virtual Private Cloud (Amazon VPC) to automatically respond to demand, adding and removing EC2 instances in addition to adjusting EC2 instance types on-the-fly.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-lyft
5:52

Chris Lambert, CIO of Lyft, explains how they built a highly available service discovery solution on AWS. Lyft’s 150 microservices serve more than 1 million requests per second and are maintained by a team of 30 who deploy code hundreds of times a day. Performance and high availability are critical. You'll learn about how they achieved their architectural goals by leveraging clusters of EC2 instances, making asynchronous calls, decoupling the components of their architecture, and storing state reliably in Amazon DynamoDB.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-sophos-architecture
6:04

Rich Vorwaller, Senior Product Manager at Sophos, explains how they dynamically scale the Sophos UTM solution to inspect network traffic of their EC2 instances using the "Elastic Load Balancing (ELB) sandwich" design pattern and Auto Scaling. You'll also learn how they leveraged Amazon S3 notifications with Amazon Simple Notification Service (Amazon SNS) to automate policy distribution to worker nodes. Lastly, you'll see how they built an outbound gateway solution using GRE tunnels and Amazon VPC networking features to provide outbound content filtering for Amazon WorkSpaces and other use cases.

View the architecture diagram.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-talend
4:12

Ashwin Viswanath, Director of Cloud Product Marketing at Talend, explains how their open source solution pulls together disparate data sources for a major pharmaceutical company to deliver real-time visibility into their business. You’ll learn how they leveraged services such as Amazon Simple Storage Service (Amazon S3) and Amazon Simple Queue Service (Amazon SQS) to bring both real-time and batch-oriented operational data into an elastic, scalable infrastructure that delivers high-quality data to an Amazon Aurora data repository. Ashwin also shows how AWS Elastic Beanstalk can process data on demand and at scale to meet huge data integration challenges.

View the architecture diagram.

To learn more about the Talend Integration Cloud, see the Talend website.

Share the YouTube video.

Adobe Flash Player or a modern browser is required to view videos on this site.

tma-palo-alto
5:39

Warby Warburton, Technical Marketing Engineering Manager at Palo Alto Networks, explains how they scale the VM-Series firewall for AWS across multiple Availability Zones. You'll learn about several advanced techniques for scaling your solutions on AWS: the "ELB sandwich" design pattern that allows you to scale different application tiers independently; using Amazon S3 to store data needed to license and configure new instances as they come online; using AWS Lambda functions to configure instances once launched; and publishing custom metrics to Amazon CloudWatch that Auto Scaling can use to accurately match capacity to demand.

Read this Tech Brief on Auto Scaling the VM-Series on AWS.

Share the YouTube video.