AWS re:Invent – what you need to know
Over the last week, we have been keeping a close eye on Las Vegas and news emerging from AWS re:Invent, the company’s annual flagship conference.
Over the five-day event, we have been treated to a plethora of new products and updates. Given the volume of news, you would be forgiven from suffering from a little overwhelm, so take five as we walk you through the key news and how it will benefit you and your organisation.
At their core, these updates reinforce AWS’s focus on enabling AI workloads, optimising costs, and improving operational reliability… let’s dive in.
Key themes from the event you need to know about
AI Factories to transform existing infrastructure for high-performance AI
Building advanced AI infrastructure demands huge upfront costs for specialist hardware, new data centres, and massive power upgrades – often leading to multi-year projects that pull focus away from core business goals. AWS has launched AWS AI Factories to transform an organisation’s existing data centre into a dedicated, high-performance AI environment. It will integrate cutting-edge components like NVIDIA GPUs and AWS networking with powerful AI services like Amazon Bedrock and SageMaker.
The benefit to organisations will be that they can leapfrog multi-year build timelines and significant capital costs by deploying a state-of-the-art AI infrastructure directly within their existing data centres, immediately satisfying data sovereignty and performance needs.
AI orchestration goes Durable and Stateful
The biggest change in cloud software is that complex, long-running processes are finally becoming stable. AWS Lambda Durable Functions will now automatically manage the progress, retries, and failures of these jobs, removing difficult manual coding when building AI services and multi-step data flows. This is paired with Bedrock AgentCore upgrades, which will add governance rules and memory across interactions, ensuring AI agents work reliably and securely before and during use.
Interestingly, Matt Garman, AWS CEO has said that AI agents will have ‘as much impact on your business as the internet or cloud’. We expect that this will take some time to take hold globally as organisations better understand how they can successfully introduce AI, but time will tell.
S3 and OpenSearch evolve into AI data foundations
Amazon S3 is transforming from a basic data repository into a cost-effective foundation for building AI services. New S3 features introduce dedicated storage that allows AI to perform search based on meaning and concepts across massive amounts of data, reducing the need for separate, expensive AI databases.
Additionally, Bedrock Knowledge Bases can now use audio and video files to enhance AI search capabilities. This change allows organisations to unlock institutional knowledge stored in videos or recordings, making that multimedia content immediately available to be queried and referenced by AI applications.
Financial management formalised with database savings
A major gap in cloud spending has been closed with the introduction of Database Savings Plans, which offer significant cost reductions when committing to use Amazon RDS and Aurora databases long-term. This is enhanced by new instance options (M7i/R7i) for RDS and SQL Server, allowing larger businesses to combine database workloads and better manage tasks that require a lot of processing power, maximising the return on expensive database software licenses.
Better security and increased operations
Core services also received upgrades focused on compliance and simplicity. Amazon GuardDuty will now provide deeper protection against threats for servers and application containers, giving teams better security visibility. For operational staff, Amazon CloudWatch Logs will work directly with S3 for simpler data analysis, and EKS added new tools to make managing large-scale containerised applications easier.
What these updates mean for you
This week’s updates reinforce AWS’s continued investment in AI infrastructure, cost optimisation, and operational excellence.
Lambda’s durable functions and Bedrock AgentCore improvements will make it significantly easier to build reliable AI applications. S3 Vectors and OpenSearch GPU acceleration will address the growing need for efficient vector storage and search. Database Savings Plans and RDS storage enhancements help organisations improve their database costs.
These capabilities offer immediate value for teams building AI applications, managing large databases, or seeking to reduce cloud spending.
If you need guidance implementing any of these features or would like a more in-depth run down of the news from AWS re:Invent, the team at CirrusHQ is here to help.
Additionally, if you aren’t already subscribed, sign up to the CirrusHQ Daily Newsletter for all the latest AWS news and updates: https://mailchi.mp/cirrushq/daily-aws-news
