🦸 Community Superhero

Our community superhero this week is Subramanya N, founding engineer at Dylog. Subramanya is a great blogger and wonderful resource for learning about AI concepts and industry trends around AI. Thank you for all your hard work and for helping teach the community, Subramanya!

💯 Spotlight

January is always a fun time of year because it’s full of bold predictions - and I love it. One prediction I found last week was from Daniel Cummins making an exciting prediction about Edge AI in 2026. Daniel lists some neat use case for SLMs (small language models) and how they fundamentally change the experience of using generative AI. He also makes a case for agentic development (aren’t we all) and for physical systems powered by AI which is as exciting as it is scary. Very cool thoughts and I think I hope he’s on the money 😅

🔥 My Favorite Content

I’ve been getting more and more into these technically deep articles that explain low-level design and how that makes an impact at an app level. Rick Houlihan published a great write-up last week explaining binary document protocols and how they are different, focusing primarily on BSON and OSON. While you can see the Oracle influence throughout the article, it has some wonderful objective analysis as Rick carefully explains the nuance of how these protocols behave. Love it!

I published an article last week detailing an update I made to Ready, Set, Cloud. I’ve been serving images directly from S3 for years, and until recently that worked fine. But with an uptick in traffic, performance started bogging down. I wrote about how I optimized images for my site and explain why you should do the same. Bonus - I set everything up in the Serverless Application Repository, so you can deploy it directly from your existing stacks with almost no extra effort!

Nabin Debnath dropped a knowledge bomb on us last week with his post on composite SLOs for event-driven systems. This article is packed with architectural goodness, anti-patterns, and hard truths. I like how Nabin approaches the problem with technical detail that follows a strong narrative arc. It’s well-written and you can tell comes from learning things the hard way.

We see a lot of articles talking about agentic AI and autonomous agents, but we don’t see very many on how to build them. Ayman Mahmoud published a piece last week called Stop building chatbots: the case for infrastructure driven AI agents. It’s a great walkthrough that shows you how to use Step Functions, Bedrock, and Lambda to build deterministic, orchestrated workflows that take advantage of LLMs. 2026 is shaping up quickly to be a year of “ok, let me build this for real,” and what Ayman is saying in this article is a contender for the “right” way to do it.

💡 Tip of the Week

I definitely live in an AI bubble. I’m always building with the newest tech or features that come out in an effort to help teach the right practices. It’s posts like this from Armand Ruiz that remind me that there’s still a lot of work to do before the greater tech community is truly ready to say “we adopted AI and do it well.”

🐣 New Releases

Reminder, all releases from AWS can be found on AWS News by Luc van Donkersgoed. Below are my favorite from last week.

Lambda now supports .NET 10. If you’re a Lambda Powertools user, this is also supported in the new runtime.

AWS Config supports 21 new resource types including CloudFront KV store and S3 TableBucketPolicy.

Last Words

Alright, first work week into 2026 is done! How was it coming back? I’ve managed to switch off vacation mode and am ready to to buckle down and build some super cool stuff. What are your plans for this year?

That’s my take on the week, but what’s yours?

What did I miss? What made you nod along (or 🙄)? Hit reply if you’re reading the email. Prefer socials? Ping me on Twitter, LinkedIn, or email.

Happy coding!

Allen