The overwhelmed person’s guide to Google Cloud: week of January 16
Richard Seroter
Chief Evangelist, Google Cloud
A weekly curation of the most helpful blogs, exciting new features, and useful events coming out of Google Cloud.
The content in this blog post was originally published last week as a members-only email to the Google Cloud Innovators community. To get this content directly in your inbox (not to mention lots of other benefits), sign up to be an Innovator today.
New and shiny
Three new things to know this week
- Trillium TPUs are generally available. These powerful processors trained Gemini 2.0, and represent a breakthrough in price performance. Whether doing model training or inference, you’ll see some fantastic benefits. And TPU Trillium-based machines are now available for use in GKE Standard or Autopilot clusters.
- AlloyDB Omni updated with key new features. I’ve got a few AlloyDB references in this newsletter edition, and won’t apologize for it. Run this powerful PostgreSQL database anywhere with AlloyDB Omni. It now features support for PostgreSQL 15.7, columnar engine on ARM, new caching capabilities, along with an updated Kubernetes operator.
- Preview of new Parameter Manager service is now available. Where are you stashing configuration settings, environment-specific settings, and secrets used by your apps? We just added an extension to Secret Manager named Parameter Manager. This gives you a centralized store for app parameters, while still getting the encryption, versioning, and regional support from Secret Manager. Check it out!
Watch this
Learn about function calling with LLMs. How do you bring real-time data and activities into your LLMs? Aja and Jason tell you all about function calling. Don’t miss their other video on preparing data for LLMs!
Community cuts
Every week I round up some of my favorite links from builders around the Google Cloud-iverse. Want to see your blog or video in the next issue? Drop Richard a line!
- Going from prototype to production with AI analysis. I like seeing posts from you all where you start with some experiments—here, in Vertex AI Studio—before turning your work into a deployable artifact. Robert does that here with some LLM prompting that turns into code deployed to Cloud Run.
- Use Firebase Data Connect for easier relational database interactions. Have you seen this in action yet? It’s pretty rad, and makes storing and retrieving data simpler. Check out this post which gives you a thorough look at how it works.
- Maybe AlloyDB is all you need. By all means, use BigQuery. But Danilo points out that AlloyDB is a viable choice for transactional and analytical workloads.
Learn and grow
- There’s still time to register for the Google Cloud x MLB™ Hackathon! Major League Baseball is calling all developers to join the ultimate challenge: build groundbreaking fan experiences on Google Cloud. Win a chance to demo your creation during the Next developer keynote and even score a trip to the 2025 MLB® All-Star Game® presented by Mastercard. Step up to the plate and register today to compete.
- Try out AlloyDB Omni’s columnar engine for analytical queries. This powerful feature of AlloyDB is also included in the run-anywhere AlloyDB Omni. Check out this codelab walks you through installing AlloyDB Omni on a Compute Engine VM and then testing out the columnar engine.
- Build a RAG-capable app using Vertex AI services. This is a new architecture guide that I found useful for understanding the role of Vertex AI and Vector Search in a generative AI app. It includes diagrams, design considerations, and much more.
- Choose between Managed Service for Apache Kafka and Pub/Sub. Both are great options for event-driven architectures. But this new guide helps you distinguish more between the two, and the ideal use cases.
- Learn how to do supervised fine-tuning of an LLM. You might know when to do SFT, but do you know how? Erwin and Bethany have a detailed but digestible post that explains the process.
One more thing
You can now “install” Google AI Studio as an app on desktop or mobile. Logan shares that Google AI Studio is now a progressive web app. I’ve got it as a dedicated app on my local machine now.
Become an Innovator to stay up-to-date on the latest news, product updates, events, and learning opportunities with Google Cloud.
