The AWS Summit London 2023 keynote was posted to AWS’s event channel shortly after the event, and it is the single best hour of content to watch if you are trying to calibrate where AWS is pushing the ecosystem. This post pulls out the session and the themes I found most useful.
Keynote
The opening keynote featured the typical mix of AWS leadership, senior AWS product voices, and UK customer stories.
Speakers:
- Tanuja Randery - Managing Director, EMEA, AWS
- Swami Sivasubramanian - VP for Databases, Analytics, AI and ML, AWS
- Renee Hunt - Chief Technology Officer, Compare the Market
- Will Cavendish - Global Digital Services Leader, Arup
Themes worth noting
Three threads ran through the keynote and most of the top-level sessions:
Generative AI moving from novelty to platform
Swami Sivasubramanian’s section is the one to watch if you want AWS’s positioning on generative AI as of mid-2023. The emphasis was on foundation model choice (via Bedrock), customer data staying inside the customer’s account, and inference-infrastructure economics via Inferentia and Trainium.
Data strategy as an AI prerequisite
The recurring message was that organisations cannot bolt generative AI onto a broken data estate. Renee Hunt’s segment on Compare the Market’s modern data platform was a concrete example of the “get the data right first” argument.
Sustainability moving from talking point to architecture decision
Will Cavendish’s contribution on behalf of Arup framed sustainability as a first-class architectural concern rather than a reporting exercise - a theme that has only strengthened in subsequent AWS events.
Other sessions worth seeking out
AWS typically publishes the full Summit session catalogue on the AWS Events YouTube channel. Sessions that tend to age well:
- The level 300 serverless and container deep-dives (EKS, Lambda, Fargate)
- Customer architecture talks, which usually hold up better than product launches
- The security track, particularly anything on Zero Trust and IAM Identity Center