|
AWS has reached its twentieth anniversary! With a gradual tempo of innovation, AWS has grown to supply over 240 complete cloud companies and continues to launch hundreds of recent options yearly for tens of millions of shoppers. Throughout this time, over 4,700 posts have been revealed on this weblog—greater than double the quantity since Jeff Barr wrote the tenth anniversary put up.
AWS modified my life
Reflecting on what I used to be doing 20 years in the past, I met Jeff in Seoul on March 13, 2006, when he got here because the keynote speaker for the Korea NGWeb convention. At the moment, Amazon was one of many first pioneers to provoke an API economic system, introducing ecommerce API companies. After the keynote speech, he returned dwelling that night, and I imagine he wrote the Amazon S3 launch weblog put up on the flight again to the USA.
That quick assembly with him introduced vital adjustments to my life. He turned my function mannequin as a blogger, and I started constructing API-based companies in my firm and opening them to third-party builders. Once I was a PhD scholar whereas taking a break from work, I spotted that for particular person researchers like me, AWS Cloud companies are highly effective instruments for conducting large-scale analysis initiatives. After returning to work, my firm turned one of many first AWS prospects in Korea in 2014. Numerous builders—myself included—have embraced cloud computing and actively used its capabilities to perform what was beforehand inconceivable.
Over the previous decade, the know-how panorama has reworked dramatically. Deep studying emerged as a breakthrough in AI, evolving by generative AI primarily based on giant language fashions (LLMs) to at the moment’s agentic AI know-how. Jeff wrote, “When trying into the longer term, you want to have the ability to distinguish between flashy distractions and real developments, whereas remaining versatile sufficient to pivot if yesterday’s area of interest turns into at the moment’s mainstream know-how.” This precept guides how AWS approaches innovation—we begin by listening to what prospects really want. The actual development isn’t pursuing each rising know-how, however moderately reimagining options that tackle prospects’ most crucial challenges.
20 years of AWS
For the primary 10 years, Jeff chosen his favourite AWS launches and weblog posts. Amazon S3, Amazon EC2 (2006), Amazon Relational Database Service, Amazon Digital Non-public Cloud (2009), Amazon DynamoDB, Amazon Redshift (2012), Amazon WorkSpaces, Amazon Kinesis (2013), AWS Lambda (2014), and AWS IoT (2015).
Whereas I additionally hate to play favorites, I need to select a few of my favourite AWS weblog posts of the previous decade.
- Deploying containers simply (2014) – Amazon Elastic Container Service makes it easy so that you can run any variety of containers throughout a managed cluster of Amazon EC2 situations utilizing highly effective APIs and different instruments. In 2017, we launched Amazon Elastic Kubernetes Service as a completely managed Kubernetes service and AWS Fargate as a serverless deployment choice.
- Excessive availability database at world scale (2017) – Amazon Aurora is a contemporary relational database service providing efficiency and excessive availability at scale. In 2018, we launched Amazon Aurora Serverless v1, and this serverless database advanced to Amazon Aurora Serverless v2 to scale all the way down to zero. In 2025, we additionally launched Amazon Aurora DSQL is the quickest serverless distributed SQL database for all the time out there purposes.
- Machine studying (ML) at your fingertips (2017) – Amazon SageMaker is a completely managed end-to-end ML service that knowledge scientists, builders, and ML consultants can use to shortly construct, prepare, and host machine studying fashions at scale. In 2024, we launched the subsequent era of Amazon SageMaker, a unified platform for knowledge, analytics, and AI and launched Amazon SageMaker AI to focus particularly on constructing, coaching, and deploying AI and ML fashions at scale.
- Greatest worth efficiency for cloud workloads (2018) – We launched Amazon EC2 A1 situations powered by the primary era of Arm-based AWS Graviton Processors designed to ship one of the best worth efficiency on your cloud workloads. Final 12 months, we previewed EC2 M9g situations powered by AWS Graviton5 processors. Over 90,000 AWS prospects have reaped the advantages of Graviton supporting fashionable AWS companies equivalent to Amazon ECS and Amazon EKS, AWS Lambda, Amazon RDS, Amazon ElastiCache, Amazon EMR, and Amazon OpenSearch Service.
- Run AWS Cloud in your knowledge heart (2019) – AWS Outposts is a household of totally managed companies delivering AWS infrastructure and companies to nearly any on-premises or edge location for a very constant hybrid expertise. Now, AWS Outposts is out there in a number of kind components, from 1U and 2U Outposts servers to 42U Outposts racks, and a number of rack deployments. Clients equivalent to DISH, Fanduel, Morningstar, Philips, and others use Outposts in workloads requiring low latency entry to on-premises programs, native knowledge processing, knowledge residency, and software migration with native system interdependencies.
- Greatest worth efficiency for ML workloads (2019) – We launched Amazon EC2 Inf1 situations powered by the primary era of AWS Inferentia chips designed to offer quick, low-latency inferencing. In 2022, we launched Amazon EC2 Trn1 situations powered by the primary era of AWS Trainium chips optimized for top efficiency AI coaching. Final 12 months, we launched Amazon EC2 Trn3 UltraServers powered by Trainium3 to ship one of the best token economics for next-generation generative AI purposes. Clients equivalent to Anthropic, Decart, poolside, Databricks, Ricoh, Karakuri, SplashMusic, and others are realizing efficiency and price advantages of Trainium-based situations and UltraServers.
- Construct your generative AI apps on AWS (2023) – Amazon Bedrock is a completely managed service that provides a selection of business main AI fashions together with a broad set of capabilities that it’s essential to construct generative AI purposes, simplifying improvement with safety, privateness, and accountable AI. Final 12 months, we launched Amazon Bedrock AgentCore, an agentic platform for constructing, deploying, and working efficient brokers securely at scale. Now, greater than 100,000 prospects worldwide select Amazon Bedrock to ship personalised experiences, automate advanced workflows, and uncover actionable insights.
- Your AI coding companion (2023) – We launched Amazon CodeWhisperer because the business’s first cloud-based AI coding assistant service. The service delivered code era from feedback, open-source code reference monitoring, and vulnerability scanning capabilities. In 2024, we rebranded the service to Amazon Q Developer and expanded its options to incorporate a chat-based assistant within the console, project-based code era, and code transformation instruments. In 2025, this service advanced into Kiro, a brand new agentic AI improvement software that brings construction to AI coding by spec-driven improvement, taking initiatives from prototype to manufacturing. Not too long ago, Kiro previewed an autonomous agent, a frontier agent that works independently on improvement duties, sustaining context and studying from each interplay.
- Broaden your AI mannequin selections (2024) – We launched Amazon Titan fashions additional rising cost-effective AI mannequin selection for textual content and multimodal wants in Amazon Bedrock. At AWS re:Invent 2024, we introduced Amazon Nova fashions that delivers frontier intelligence and business main worth efficiency. Now Amazon Nova has a portfolio of AI choices—together with Amazon Nova fashions, Amazon Nova Forge, a brand new service to construct your personal frontier fashions; and Amazon Nova Act, a brand new service to construct brokers that automate browser-based UI workflows powered by a customized Amazon Nova 2 Lite mannequin.
Construct with AI: Your path ahead
A decade in the past, AWS responded to the emergence of deep studying by launching the broadest and deepest ML companies, equivalent to Amazon SageMaker, democratizing AI for a variety of shoppers—from particular person builders and startups to giant enterprises—no matter their technical experience.
AI know-how has superior considerably, however constructing and deploying AI fashions and purposes nonetheless stays advanced for a lot of builders and organizations. AWS affords the broadest number of AI fashions by Amazon Bedrock, together with main suppliers equivalent to Anthropic and OpenAI. By utilizing our mannequin coaching and inference infrastructure and accountable AI each sensible and scalable, you’ll be able to speed up trusted AI innovation whereas sustaining management of your knowledge and prices—all constructed on our world infrastructure’s operational excellence.
Reinvent your concept, carry on studying, construct confidently with AI you’ll be able to belief, and share your successes with us! New AWS prospects obtain as much as $200 in credit to attempt AWS AI free of charge. For those who’re a scholar, begin constructing with Kiro free of charge utilizing 1,000 credit per 30 days for one 12 months.
— Channy