1 month ago
9
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
- Create and manage your security posture from within the Cloud Console. Let’s focus on security this week. This first product update is about a new feature of Security Command Center where you can create security postures—it comes with pre-defined templates—that are applied to your whole environment. Powerful feature!
- Cloud Run updated with new security features. First, you can now disable the default URL for a Cloud Run service to ensure you only allow traffic from explicitly allowed sources. And for deployed-from-source scenarios, we now provide automatic base image upgrades. This means that you get OS, OS package, and language runtime patches without needing a new deployment or incurring downtime.
- Cloud KMS with Autokey makes encryption key management easier. Provision and assign encryption keys automatically with this new feature. It works with Cloud Storage, Compute Engine, BigQuery, Secret Manager, Cloud SQL, and Spanner. Read the new docs and watch this video overview.
- Open ML Platforms: Technical Challenges & Solutions. Discover how Open ML platforms can optimize collaboration, streamline resources, and drive innovation in AI. Join Google Cloud experts as they explore solutions to tackle technical challenges and enhance your ML operations.
Watch this
A developer’s guide to LLMs. By developers, for developers. Aja and Jason kick off a series of videos with a look at use cases for generative AI, how to pick a model, cost considerations, and even a demo.
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!
- How to build data pipelines that feed BigQuery ML. Jana offers up a detailed walkthrough for using Cloud Storage, Apache Airflow, BigQuery, and more to get insights from her data.
- Get your cloud logging set up with Terraform. Establish a repeatable logging approach using infrastructure-as-code. Aaron takes a look at how you’d make that happen.
- Store, serve, and monitor ML models like a pro. I liked Maxwell’s post which explains some of the core tasks of an ML Engineer. See what it takes to get an ML model deployed for online prediction in Vertex AI.
Learn and grow
Three ways to build your cloud muscles this week
- Read from a shared SQLite database from Cloud Run. It’s just a simple example from Steren, but I liked the pattern. Store a SQLite database in Cloud Storage, and mount that as a volume to Cloud Run instances for a cheap, fast, and easy database.
- It’s time to upgrade your platform. What’s a modern platform look like? I wrote a paper to answer that question, and figure out what’s needed to accommodate new AI workloads. Enjoy!
- Plan for business continuity in your CI/CD pipelines. I haven’t seen much written on this topic. This is a good look at how you keep your software delivery process running even when the unexpected happens.
- Four ways to get better responses from your LLM. We just published this paper which explores four relevant techniques—prompt engineering, RAG, fine-tuning, long-context windows—that you can use to shape the results of an LLM call.
One more thing
New Gemini 1.5 Flash brings great performance at a low price. It’s hard (impossible?) to beat the price performance of our Gemini 1.5 model family, and now you can do some remarkable things with Gemini 1.5 Flash for peanuts.
Become an Innovator to stay up-to-date on the latest news, product updates, events, and learning opportunities with Google Cloud.
Posted in