#28 Faster Serverless! 🏃🏽‍♀️‍➡️

This week, our expert is AWS Community Builder Christian Bonzelet, our spotlight falls on AWS Serverless Hero Angela Timofte, we look at the latest AWS service releases, blog posts, & more!

Mar 14, 2025

Welcome

In last week’s issue, our serverless expert was AWS Community Builder Sebastian Bille, and our spotlight fell on AWS Principal Developer Advocate Marcia Villalba!
This week, our serverless expert is AWS Community Builder Christian Bonzelet, our spotlight falls on AWS Serverless Hero Angela Timofte, we look at the latest AWS service releases, blog posts, hints and tips, news and more!
This week’s newsletter is sponsored by Leighton.

📰 Articles that caught the eye

Here are some of the stand-out articles I read during the week in the World of Serverless.

⭐ My favourite this week was the article by Jason Conway-Williams on Internationalisation as a software pattern.

🎓 Ask the Expert

Each week, I ask a different serverless expert the same three questions to get their personal insights - this week, we have AWS Community Builder Christian Bonzelet!

1. What is one common mistake you see teams making when implementing serverless solutions, and how can they avoid it?

In my role as a Solution Architect (aka 'a Trusted Advisor'), I am trained in not speaking about mistakes. However, I see some recurring misconceptions that I typically challenge in my role. One of the most prominent is a lack of analysing trade-offs on how to tackle a problem, either using a managed (serverless) service or by building the solution on your own.

I observe that decisions are often HIPPO-driven—HIPPO standing for Highest Paid Person's Opinion—a dangerous anti-pattern in decision-making. I encourage teams to avoid this by conducting a thorough trade-off analysis.

The process of documenting architecture decisions using Architecture Decision Records (ADRs) can be a good starting point, as it helps you not only document the decision but also explain, discuss, and evaluate all options and their consequences.

Having to specify the impact of an architecture decision forces teams to think about whether those impacts outweigh the benefits. For example, when choosing between encoding videos via ffmpeg on EC2 instances, ECS/Fargate or using AWS Elemental MediaConvert to do the job, teams should consider factors like operational overhead, scalability needs, and cost implications rather than defaulting to familiar approaches.

2. Which serverless tool or service are you most excited about right now, and why?

I've always been, and remain excited about the broad range of serverless AI services that AWS offers, especially things like Amazon Personalize, Amazon Translate, Amazon Rekognition or Amazon Comprehend.

Those services make it so easy to integrate AI capabilities into your applications and workflows, often with just a single API call. Another great set of features that I am very excited about is the announcements of supporting private API integrations in Amazon Eventbridge and Step Functions.

I work in a hybrid ecosystem of solutions built for the cloud and solutions running on-premise. Making these two worlds talk to each other while limiting complexity to build up the required networking infrastructure ultimately simplifies our architectures.

3. What is your favourite trick or tip when working with serverless that the readers may find interesting?

Imagine you are designing a solution with a team in front of a Whiteboard. It all looks simple at first but gets more and more complicated the deeper you dig into the business domain. You and your team start drawing an Amazon API Gateway, some AWS Step Function state machines and an Amazon DynamoDb Table. You end up with more integrations to Amazon S3, Amazon SQS, Amazon SNS or Amazon EventBridge. You might ask yourself, why and when did this start to get complicated?

I use the C4 model to design software solutions. This model provides several layers (Context, Containers, Components, and Code) with defined purposes and perspectives of your solution. Every layer is connected and allows you to zoom in and out at will, giving you flexibility in how you visualise and communicate your architecture.

Whenever I get this overwhelming feeling of too much complexity or whenever I need too much time to process the amount of information transported in a diagram - I know it is time for a new perspective. It is time to step back, zoom out and create a context diagram providing a holistic view of the system we are going to build or expand.

This helps me to shift the focus from just looking at the technical-centric status quo, more into a customer-centric state-to-be. Zooming out helps me to embrace a problem-first thinking. With the system context in your mind, you zoom back into your solution diagram and can make better-informed decisions about the essential parts of your solution.

✅ Bonus tip: join the hashtag#believeinsls discord! There is a community there to answer any questions you may have without getting overzealous on serverless or without judgment! Check it out!

🚀 New Releases

Here are the latest releases this week in the AWS World:

⭐ The most interesting for me this week is the multi-agent collaboration on Amazon Bedrock!

🔥 Tip: Check out https://aws-news.com/ for the very latest up-to-date serverless releases as they happen, created by the talented AWS Serverless Hero Luc van Donkersgoed.

👷🏻 Tools & Frameworks

Check out the latest open-source frameworks, news, and tool updates from the past week.

TypeScript - A 10x Faster TypeScript.
s3-migrate - A command-line tool to migrate objects from one S3-compatible storage bucket to another. Supports resuming interrupted transfers using a SQLite state file.
opentelemetry-lambda - OpenTelemetry Lambda Node.js layer.
ServerlessLand - Accelerate serverless development with ready-to-use Serverless Land Patterns in Visual Studio Code.

✖️ Social of the Week

This week’s social is on X/Twitter by Haider, which has a video of Anthropic CEO Dario Amodei discussing that he feels in 12 months’ time, almost all code that is written will be generated using AI.

The reason I find this super interesting is some people will see this as scaremongering and marketing, and be concerned for jobs (heads in the sand), whilst some others will (or have already) embraced AI code generation and will see this as a new extension to how they work daily.

Do you think in 12 months, 90% of code will be generated using AI?

🎙️ YouTube & Podcasts

Here are some of my favourite videos and podcasts this week.

⭐ My favourite this week was the video by Yan and Khawaja on building multi-tenant systems!

💡 Hints & Tips

Each week, I share quick hints or tips based on things I notice in day-to-day engineering life. This week we will focus on Amazon Cognito.

⭐ [Tip 1] Cognito Custom Domains - this week we are going to discuss a ‘quirk’ with using custom domains with Amazon Cognito UserPools that can be problematic for teams when you hit it for the first time.

One issue people hit is that you need to create a “dummy” A record for the parent of your sub domain if there isn’t already a valid one, or CloudFormation will fail the deployment with a not so helpful error message! For example, with auth.app.serverlessadvocate.com we need to create a dummy A record for app.serverlessadvocate.com that points to an address like 1.1.1.1.

”You must create an A record for the parent domain in your DNS configuration. When the parent domain resolves to a valid A record, Amazon Cognito doesn't perform additional verifications. If the parent domain doesn't point to a real IP address, then put a dummy IP address in your DNS configuration, such as "8.8.8.8" - https://repost.aws/knowledge-center/cognito-custom-domain-errors

⭐ [Tip 2] Ephemeral vs long-lived - The second tip when working with ephemeral environments is to use Cognito Prefixed domains for speed when it comes to ephemeral environments (short-lived feature deployments), and Cognito custom domains for long-lived (develop, staging and production). The reason for this is the custom domains take a lot longer to deploy, so this can speed up your temporary deployments (but still giving you a deterministic auth URL).

I will write this up with a full code example and article in the coming weeks!

Weekly Case Study 🔍

This week’s case study comes from OLX.

OLX, a global online classifieds leader serving 11 million customers monthly, enhanced its customer identity and access management (hCIAM) security using AWS WAF Bot Control, AWS Shield Advanced, and Amazon Cognito to block malicious bot traffic while reducing infrastructure costs.

Running 2 million daily requests, OLX saw a bot attack spike to 2.1 million requests, rejecting 250,000 false ones and saving thousands in SMS costs. The company’s backend is powered by AWS Lambda, with AWS Fargate and Amazon ECS handling the front end, all secured by AWS WAF and AWS Shield.

Go check out the case study for more information.

🗣️ Inspirational Quotes and Thoughts

This week’s inspirational quote comes from Aristotle:

"The purpose of knowledge is action, not knowledge"
- Aristotle

This quote came to mind during the week when discussing the value of AWS Certifications with a group of people at work (as many of you may know, I am very much into stoicism!), and the reason was that I feel a lot of people accumulate knowledge simply to pass the exams, yet many do this without putting the knowledge into action and getting hands-on experience.

If you are currently studying for AWS certifications, I would really encourage you to get your hands dirty (even just in the console) and try the services and patterns out first-hand. Don’t just remember the information to simply get a certificate.

🗳️ Poll of the Week

In last week’s poll, we asked the question, “What is your typical drink whilst coding your solutions?“.

50% of people said coffee, 20% Diet Coke, 20% tea, and 10% water! The 50% is no surprise, of course!

This week, we ask the question, “Do you think an architect as a role is required in a serverless team?“. Do we need a dedicated architect in the team, or is it a shared responsibility across a cross-functional team? Should it be done by a Principal Engineer? Or is it a role that is always needed within all serverless teams?

📅 Serverless Events

The following serverless events are upcoming, so mark your calendars.

🎟️ To note, CFP is currently open for the AWS North conference which I am helping organise, and we also have opportunities for lightening talks throughout the day. Go check it out!

Other fantastic events happening soon:

AWS Community Day Slovakia - 26th March 2025
AWS Community Day Italy - 2nd April 2025
AWS Community Day Romania - 10th April 2025
AWS Community Day Turkey - 19th April 2025
ACD Bengaluru - 23rd May 2025
ACD Adria - 5th Sept 2025
AWS Community Day Baltic - 10th Sept 2025
ACD DACH - 7th Oct 2025
AWS North Community Conference - 16th Oct 2025

Do you have any upcoming events that you want to highlight? Message me below!

⭐ Spotlight

This week’s spotlight falls on AWS Serverless Hero Angela Timofte:

Angela is an AWS Serverless Hero and VP of Global Engineering at Trustpilot, recognised for her fantastic leadership and vast experience in migrating legacy applications to modern, serverless, event-driven architectures.

As a co-organiser of the Copenhagen AWS User Group and a seasoned frequent international speaker at AWS events (such as Serverless Days Manchester, where we both spoke recently, AWS re:Invent, and many, many keynote engagements), Angela is dedicated to knowledge sharing and empowering others in the serverless community. Many thanks for all you do for our fantastic global community, Angela!

Thank you for reading the latest Serverless Advocate Newsletter!

If you want to find out a little more about me, please have a look at:

https://www.serverlessadvocate.com/

See you next time,

Lee