I’m happy to wrapping a long, productive week. I’m hoping to have a bit of tech play time this weekend, and have a few projects in mind. I hope you find a way to enjoy yourself and do something fun.
[blog] In-Place LLM Insights: BigQuery & Gemini for Structured & Unstructured Data Analytics. Abi covers a lot of ground in this piece, and shows us both inline model calls, as well as invoking remote functions that call LLMs.
[blog] In Defense of Craft. Simple, but powerful point here. Most all of work in settings where a job-well-done is based on external factors, but this post talks about the satisfaction we get from doing something that stands on its own.
[blog] Introducing Cloud Run volume mounts: connect your app to Cloud Storage or NFS. Contorting your architecture to use fully managed compute services while also banking on persistent data storage? Now Cloud Run lets you set up volume mounts to shared data sources.
[article] How People Are Really Using GenAI. This study surfaced a hundred different ways people are using generative AI technology. There aren’t a ton of surprises here, but it’s helpful to see it listed out.
[blog] Workload Identity – Secured Way to Access Google Cloud APIs from GKE Workloads. This has thankfully become a common idea for Kubernetes clusters, so make yourself familiar with it. Workload identity gives you a more secure way to control access from a workload to a given cloud resource (database, etc).
[article] Java 22: Making Java More Attractive for AI Apps/Workloads. It doesn’t look like any language supporters are willing to cede the AI territory to Python. Lots of languages are beefing up their readiness and applicability for AI workloads.
[blog] Modeling Extremely Large Images with xT. The folks at Berkeley are proposing a new framework to understand very large images. Very cool.
[docs] Work with Salesforce Data Cloud data in BigQuery. Wow, this seems like a big deal. Natively query your Salesforce data from BigQuery and do cross-cloud analytics. Blog post about it here.
[blog] How to Learn and Practice Product Management in a Feature Factory. Stuck in a feature factory that pumps out new capabilities and moves on? Here’s advice on how to grow as a product manager in that setting.
[blog] Shipping quality software in hostile environments. This feels related to the above article on feature factories. When you work at places that pile on tech debt because of lack of autonomy and endless shipping, you’re in trouble!
##
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 