Today’s reading list seemed to be skewed towards “struggles.” Struggles with career progression, scaling AI programs, software supply chain risks, upgrading core infrastructure, or even doing test-driven development right. The struggle is real, but all you can do is find your next right step, and get access to those who can help you make it through.
[article] 4 Experiments to Encourage Employees’ Career Progress. Getting promoted isn’t the way we should measure career progression. There are lots of ways to advance professionally, and this article explores a few ways to think about it.
[article] Uncovering the Seams in Mainframes for Incremental Modernisation. This is the type of architecture work I can geek out over. It’s real forensics into a legacy system and finding seams that help you start incremental modernization.
[blog] What’s in a Name? Here’s a good little post with very useful advice on naming things in your codebase.
[blog] LLM&FinOps: Cost Optimization Options to Run High Performance AI/ML Workloads on GKE in Google Cloud. It wasn’t cheap for cloud providers to buy all this AI-ready hardware, and it’s not necessarily cheap for you to use it. But you can manage costs if you come into it prepared.
[article] Accenture sees companies struggling to scale AI. I found a few interesting nuggets here. Big firms (cloud vendors, consulting shops) have invested in AI ahead of revenues. But the client projects are now coming. And modernization is table stakes to take advantage.
[article] Google Cloud CISO contrasts shared fate vs. shared responsibility models. It’s an important distinction, and this article calls out what “shared fate” looks like.
[article] Tech giants grapple with ballooning software supply chain risk, JFrog report reveals. I mean, yeah, you’re going to have problems if you’re using a dozen different programming languages. Time to go on a purge, my friend.
[article] Why Isn’t the World Upgrading Its Databases? From what I’ve observed, most companies don’t upgrade their infrastructure—databases, operating systems, business apps, commercial software, programming languages—until there’s a forcing function. If it just works, even if it’s out of mainstream support, it’ll keep running until something requires a change.
[blog] The Product Model at Amazon. Read this to learn a bit about how Amazon figures out which problems to try and solve, and how they use product discovery techniques to test ideas.
[blog] Why the Frontend Kingmaker isn’t Full-Stack: A History. Be proud of being a frontend engineer! That’s tough work. Kate looks at the evolution of “acceptance” for this role.
[blog] TDD: You’re Probably Doing It Just Fine. Some pragmatic advice here for those who might panic that they’re not doing test-driven development correctly.
##
Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below:
Richard Seroter's Architecture Musings 
One thought