SaaS at re:Invent 2024
Complete list of all the SaaS sessions at re:Invent 2024.
Bill Tarr (AWS)
Amazon Employee
Published Sep 6, 2024
Last Modified Oct 4, 2024
- Make sure you get to the re:Invent catalog and mark your favorites. That was you can quickly add them when we open for full registration!
Abstract: The SaaS Builder Toolkit (SBT) provides developers with a pre-built set of tools that decompose SaaS into a series of building blocks that can be used to create multi-tenant environments. This session digs into the moving parts of this toolkit, exploring the inner working of it core components, architecture, and its extensibility model. We'll also look at real-world example of SBT in action, composing a working multi-tenant application from scratch. As part of this exploration, we'll look at how SBT addresses core concepts, including building a control plane, onboarding tenants, authenticating tenants, supporting tiering, and provisioning tenant resources.
Abstract: Multi-tenancy adds a set new architecture, design, and implementation considerations to the generative AI domain. This session will dig into a working multi-tenant generative AI solution, looking closely at how tenancy impacts the implementation of data partitioning, tenant isolation, tenant onboarding, noisy neighbor, tiering, and so on. This includes looking at how you can support per-tenant generative AI experiences that allow you to offer differentiated experience to individual tenants. The goal here is to provide an in-depth view of all the moving parts of an end-to-end SaaS experience that uses Amazon Bedrock, Knowledge bases, vector databases, and so on.
Abstract: The last 7+ years spent helping companies build SaaS solutions has been eye-opening. It's given us great insights into the dynamics, challenges, and pitfalls that teams often face when building SaaS solutions. In this session, we’ll go through a range of different patterns we’ve seen, highlighting common technical and business themes that have impacted the scale, growth, and cost efficiency of SaaS offerings. This is about capturing these trends and outlining guidance that can help teams avoid falling into these same traps. We’ll get into the technical nuances, architecture challenges, and operational impacts that undermine the success of SaaS businesses.
Abstract: SaaS builders must balance a range of competing needs when selecting a multi-tenant storage strategy. How you'll scale, how you'll prevent noisy neighbor, how you'll isolate tenant data, how you'll support remote data, how you'll shard workloads—these are amongst the many factors you must consider. This session will examine the different strategies and patterns that are applied when designing a scalable multi-tenant data architecture that can evolve as your business evolves. This goes beyond data partitioning, looking into the details of how you address challenging multi-tenant storage models, balancing cost, operational, and performance efficiency.
Abstract: SaaS providers often need to support of authentication and authorization models, allowing tenants to access their system through internally and externally identity sources. This session will look at how you can build a federated multi-tenant identity experience that could be built in different combinations with Amazon Cognito, AutoO, Ping, Okta, and so on. We’ll dig into the implementation and architecture details, looking at how the nuances of these different identity providers will shape the onboarding, tenant context injection, and tenant user management lifecycle for your SaaS environment.
Abstract: Building a multi-tenant SaaS architecture is a balancing act. Finding the right blend of cost efficiency, scale, resilience, and operational agility can be challenging. It’s here that a cell-based architecture can often represents a natural fit. The cell-based model gives you new ways to approach tenant isolation, deployment, scaling, regional distribution, and tiering, bringing a range of new possibilities to your SaaS architecture and operational footprint. In this session, we’ll go under the hood of a cell-based multi-tenant architecture, highlighting the strategies, patterns, and considerations that come with employing this model.
Abstract: Multi-tenant SaaS environments, require builders to address a challenging set of of scaling, resilience, automation, and operational efficiency challenges. There a good tools in this space, but it’s time to ask: what’s next? In this session, we’ll outline new multi-tenant architecture, DevOps, and operational strategies that challenge the current boundaries/models. This will include a deeper look at the data you need to collect, the insights you should be gathering, and to emerging tools/mechanism that can enrich this experience. We’ll explore new approaches to multi-tenant operations, looking at how generative AI can enrich your view into the efficacy of your scaling, resilience, isolation, and tiering policies as well as its potential to impact the operational profile of your SaaS business.
Abstract: Amazon EKS offers SaaS builders a diverse range of tools and constructs that can be used to support the complex needs of multi-tenant environments. Navigating the landscape of EKS options can be challenging. This session will explore the various EKS tools, techniques, and strategies that can be used to address core multi-tenant requirements, including isolation, deployment, tiering, resilience, and cost. As part of this, we'll look at how Karpenter, native AWS services, and Kubernetes community tools can be used to shape the scale, efficiency, and cost profile of your SaaS offering.
Abstract: The path to SaaS is different for many organizations. There are a range of different business and technical factors that will influence how you choose to move an existing application to a multi-tenant delivery model. This session will look at a range of SaaS modernization strategies and patterns, exploring the challenges and tradeoffs associated with different approaches. How/where/when to do you make this move without requiring a complete rewrite? How do you inject tenant context into your experience? What parts of the SaaS control plane are needed? These are samples of the questions that we'll cover in this session.
Abstract: Multi-tenancy has a distinct and substantial influence on the DevOps footprint of SaaS environments. The unique deployment, partitioning, tiering, isolation, and onboarding profile of SaaS solutions require teams to embed support for these concepts directly into their DevOps automation. This session will look different tools, strategies, and patterns that are commonly used in multi-tenant environments to create a deployment and onboarding experience that can support the blend of architecture requirements imposed by different tenant personas. We'll use this setting to talk through the myriad of options you have here, mapping these strategies to different multi-tenant realities/considerations.
Abstract: SaaS organizations are often driven by growth. Scaling to meet this growth often requires teams to think about how their underlying architecture, operations, and application can support these growth model. Designing a multi-tenant environment that can scale into new geographies, be deployed into more regions, and/or address additional compliance requirements can be challenging. This chalk talk will examine the architectural challenges that come with supporting various growth/reach models, highlighting techniques, patterns, and strategies that are used to prime your SaaS offering for broader reach/expansion. It covers the architectural, deployment, resilience, and operational considerations that come with tackling this growth profile.
Abstract: There is no one-size-fits-all SaaS architecture. Instead, architects must find the unique mix of multi-tenant strategies that address the needs of their specific application. This session outlines the core SaaS patterns that architects can use to compose their multi-tenant architectures. Which isolation and deployment models best meet the needs of your solution? What strategies to orchestrate the onboarding and provisioning of tenants? These are all examples of areas where you'll need to determine which tools, services, technologies, and strategies you should have on your radar when designing a SaaS solution.
Abstract: Every SaaS provider must have a tenant isolation strategy. This is especially challenging for organizations that are trying to maximize cost efficiency by relying on shared tenant infrastructure. This session will go beyond the basic isolation schemes, looking at new technologies and techniques that can be used to isolate tenant resources. We'll explore a range of different isolation models, including RBAC, ABAC, dynamic policies, fine-grained authorization, and Amazon Verified Permissions. Attendees will learn how to leverage these techniques to maintain security and access control of their multi-tenant environments, while minimizing the operational overhead.
Abstract: SaaS applications are often required to support a variety of personas/tiers. Supporting these varied experience requires a more distributed multi-tenant architecture model where tenants/tiers may have dedicated resources. This session explores different architecture patterns, technologies, and strategies that are used to route requests to per-tenant/tier infrastructure. We'll cover a wider range of routing considerations, including domain and data driven routing, and how connectivity options like Route53, CloudFront, APIs and PrivateLink affect your routing strategy. The goal is to review the strengths and limitations of each option, helping you identify the strategy that best addresses the needs of your solution.
This chalk talk explores advanced strategies for managing multi-tenant Amazon SQS queues, discusses the challenges posed by noisy neighbors, and shares effective mitigation techniques, including shuffle sharding and overflow queues. Gain insights into optimizing queue performance, ensuring fair resource allocation, and maintaining service quality across tenants. Walk through best practices for implementing these solutions, potential trade-offs, and examples of multi-tenant Amazon SQS architectures.
Abstract: Teams are continually looking for better ways to address the unique scaling, resilience, deployment, and tiering requirements of multi-tenant SaaS environments. It can be difficult to strike a balance between complexity, efficiency, and flexibility. In this workshop, we’ll look at how cell-based architectures provide you with new ways to group, deploy, scale, and operate your multi-tenant workloads. We’ll go under the hood of a working SaaS cell-based architecture, looking at how you can implement provisioning, configuration, routing, and dynamic sharding of tenants. We’ll also look at how this approach influences the tiering, scaling, and resilience profile of your SaaS architecture. The goal here is to give you a look at the architecture considerations the come with multi-tenant cell-based model, giving you new tools and strategies for addressing the unique needs of SaaS environments.
Abstract: Building a multi-tenant SaaS offering is just part of the job. Once your environment is loaded up with tenants, you’ll still need rich, proactive tooling and mechanisms that allow you to analyze, troubleshoot, and evaluate the state of your system. This workshop will take you through a series of simulated operational events (noisy neighbor, isolation violations, etc.), creating the tools you’ll need to track and analyze the activity of individual tenants/tiers. The goal here is to highlight the operational challenges that can surface in multi-tenant environments and explore approaches to creating the tenant-aware operational mechanisms that can support this experience.
Abstract: Cost-per-tenant is an essential metric to SaaS businesses. It is the one piece of data that tells you how you architecture, design, tiering, and scaling strategies are optimizing the costs associated with individual tenants workloads. In this workshop, we’ll explore the complexities and strategies that come with designing and implementing a cost-per-tenant model.. We’ll explore the various instrumentation and consumption attribution strategies that can be applied across the layers of your SaaS architecture. The goal here is to give a clear view into the strategies and patterns that can be used to build your own cost-per-tenant model.
Abstract: Generative AI will clearly have significant impact on the footprint and experience of SaaS applications. The question is: how does generative AI influence the multi-tenant design and architecture of your SaaS architecture? In this workshop, we’ll look at how generative AI can be used to deliver all new experiences. This includes building out tenant-specific RAG experiences, configuring and managing per-tenant RAG data, metering tenant consumption of models, implementing tenant isolation with generative AI constructs, and so on. The goal here is to get into guide you through an end-to-end design and development of a working multi-tenant generative AI solution.
Abstract: SaaS Builder Toolkit (SBT) is a library of developer tools that can be combined, configured, and extended with the goal of accelerating your ability to create, modernize, or enhance the multi-tenant footprint of a SaaS solution. This workshop digs into SBT, using it to build out the core moving parts of a working SaaS environment. We’ll look at how builders can configure the various parts of a SaaS architecture, including a dive into the control plane, integration with the application plane, and the spin-up of a working application. This developer-centric journey will expose you to the end-to-end elements of SBT.
Abstract: The Kubernetes DevOps tool chain equips builders with a rich collection of constructs that can be used to design and implement a range of multi-tenant automation strategies. This workshop looks at how these Kubernetes tools are combined to support tenant onboarding, deployment, provisioning, configuration, and so on. The labs will identify approaches for targeting and automating these unique SaaS DevOps needs. The goal here is to create an end-to-end solution that uses GitOps, Argo, Helm, and other Kubernetes tools to create and configure multi-tenants EKS environments that cover a range of tiers and deployment models.
Abstract: Microservice represent the area of your SaaS application where multi-tenant context meets the everyday coding needs of SaaS developers. The microservices of your SaaS environment are where you'll ultimately apply or leverage the constructs that will address tenant isolation, accessing data with tenant context, recording tenant-aware metrics, integrating with billing systems, interacting with other multi-tenant microservices, and long list of other considerations. In this workshop, we'll create real-world microservices that cover a broad spectrum of multi-tenant use cases and implementation strategies. The goal here is to see first-hand the way SaaS influences the inner workings of your microservices.
Abstract: Teams are continually looking for better ways to address the unique scaling, resilience, deployment, and tiering requirements of multi-tenant SaaS environments. It can be difficult to strike a balance between complexity, efficiency, and flexibility. In this workshop, we’ll look at how cell-based architectures provide you with new ways to group, deploy, scale, and operate your multi-tenant workloads. We’ll go under the hood of a working SaaS cell-based architecture, looking at how you can implement provisioning, configuration, routing, and dynamic sharding of tenants. We’ll also look at how this approach influences the tiering, scaling, and resilience profile of your SaaS architecture. The goal here is to give you a look at the architecture considerations the come with multi-tenant cell-based model, giving you new tools and strategies for addressing the unique needs of SaaS environments.
Abstract: Building a multi-tenant SaaS architecture will require teams to consider wide range of design and architecture strategies. The footprint of your control plane, the services it will host, how you’ll handle onboarding, what tiers you’ll support, how you’ll isolate tenants—these are all areas you’ll need to address when building out your SaaS architecture. This session will take an end-to-end walk through building out some of these core constructs, highlighting the design and implementation decisions that will impact your approach. The goal is to give you an accelerated, in-depth look at the core elements of a SaaS architecture.
Abstract: SaaS developers and architects face a broad range of scaling, resiliency, cost, and operational efficiency challenges. As part of this effort, they also need to design and build the mechanisms that can validate their architecture assumptions, simulating loads, tenant personas, isolation exceptions, onboarding spikes, and so on. This session will take you through the process of stressing a pre-built SaaS environment, looking at how/where/when you can introduce the mechanisms that exercise and validate the policies employed by your multi-tenant architecture. Getting this right is essential to building a SaaS environment that achieves your broader operational and efficiency goals.
Any opinions in this post are those of the individual author and may not reflect the opinions of AWS.