🦸 Community Superhero

Our community superhero this week is Misbah Syed, founder at Menlo Park Labs. Misbah consistently contributes thoughtful, grounded perspectives on AI and offers practical advice for building with it. He makes great videos that are short and information dense to keep you up to date on the latest and greatest. Thank you for everything you do, Misbah!

💯 Spotlight

We’re starting off this week with a genuine reality check. One that comes from a place of practiced reasoning and an expert in looking forward. Marc Brooker published an article last week on the success of natural language programming, which is essentially his take on vibe coding and more importantly, spec-driven development. Marc poses that using natural language to build software is what we’ve always done… literally. Requirements come in the form of natural language from clients, product owners, and other stakeholders, and then we (engineers) take those and turn them into patterns and code. With LLMs, it all shifts left. Now the engineers are defining how the software should be built and AI is turning it into patterns and code. This is a grounding article that makes me feel reassured about the way our industry is trending right now.

🔥 My Favorite Content

I finally started trying out Lambda Durable Functions, but not until I read Andres Moreno’s article about them. His post likens them to Step Functions, but mostly goes over what you can do with them. He defines the important terms used in Durable Functions and shows you how to build with each of the new pieces available in Lambda. This is pretty cool and definitely makes it easier to build!

Julia Furst Morgado is a creative genius. Last week she started a “30 days of OTel” on LinkedIn and has been absolutely killing it with her clever videos and graphics. Every day last week had a captivating piece of content explaining a little nugget about observability and telemetry. Since are all less than 30 seconds, I’ll share them all. Monday, Tuesday, Wednesday, Thursday, and Friday. Be sure to follow Julia as she works her way through the rest of the 30 days!

On the opposite end of the spectrum, I stumbled upon a formal whitepaper last week about cold-start anti-patterns and refactorings in serverless systems. It’s not so much about “rediscovering” cold starts as it is pointing out dozens of real-world incidents as a case for establishing concrete anti-patterns. The big takeaway for me on this was the benchmarking suite and analysis approach. They put together a means for measuring cold starts objectively and scientifically, which is a bit of a departure from the anecdotal blogs we usually read (not in a bad way).

I am fascinated at how tech is “catching up” with AI and agentic development. It feels like capabilities skyrocketed this year but nobody knew what to do with them. Now that we’re getting a handle on everything, content is popping up around tangental AI topics. I found an article last week titled Your IT stack wasn’t built for agentic AI by Kaladhar Voruganti and Kevin Egan that discussed at length how infrastructure wasn’t really ready for the agentic boom. It is a vendor pitch article, but the arguments they’re making do stand true in most cases. I really like seeing their mental model to the physical components to AI rather than the build tools.

💡 Tip of the Week

I found a double whammy of goodness on LinkedIn last week from Andrew Zigler. He shared a little Dr. Seuss-style picture book he made with AI about how MCP works - which is brilliant in its own right. But he also shared practical advice for content creators everywhere in the copy of his post. If you want to make something stick, make it fun and memorable. Be unique. What can you do that other are not doing? I love, love, love this post!

🐣 New Releases

These are a few of my favorite releases from last week, for a full list of everything, be sure to check out AWS News.

Lambda Durable Functions made its way into 14 additional AWS regions last week. The rollout is almost complete!

You can now do cost allocation based on IAM user attributes. Pretty neat and useful feature.

DynamoDB, Aurora DSQL, and Aurora PostgreSQL are available in the Vercel Marketplace now. I like this cohesion they have going on. Plus you get $100 in AWS credits when you create an AWS account from Vercel!

Last Words

As the year winds down I’m seeing lots of 2025 recaps. What are you most proud of this year? I love reading these and seeing what motivated the community to share with others. Please share 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