- Odyssey Platform Weekly
- Posts
- Odyssey Platform # 14
Odyssey Platform # 14
Another exciting week in the Platform Engineering ecosystem!

Editor's Note
Welcome to another edition of Odyssey Platform Weekly! This week, we’re diving into fresh insights, key events, and powerful stories shaping the future of platform engineering.🗞️ In this newsletter
🗓️ Events
🔦 Tool Spotlight
🔍️ Deep Dive: Installing Crossplane providers
🎯 Stay Inspired - Case Studies
👀 In Case you missed it
Latest news & events in the platform engineering domain
📆 Upcoming Events
CloudCon Sydney 2025 is back on September 9-10, bigger, better, and powered by the community. Over two full days, cloud builders, engineers, and innovators will come together to share ideas, learn new skills, and explore the future of cloud computing, platform engineering, DevOps, and AI-powered infrastructure. Look forward to inspiring talks, hands-on workshops, and countless opportunities to connect with peers who are shaping the next generation of scalable, secure, and sustainable platforms.
👉 Register here

🔦 Tool Spotlight
Once known primarily as a lightning-fast cache, Redis has grown into a versatile data platform for real-time applications. Its in-memory architecture delivers microsecond response times, making it a natural fit for caching, session management, message brokering, and streaming workloads.
Redis’s popularity comes from its simplicity and speed. Developers can store and query key-value pairs, lists, sets, streams, and JSON documents without complex schemas, while its built-in replication and clustering enable high availability and horizontal scalability. With Redis Enterprise, teams also get support for multi-model data, automatic failover, and active-active geo-distribution.
The platform has gone cloud-native with fully managed Redis services available on AWS, Azure, and GCP, making it easy to integrate with modern architectures. Features like Redis Modules extend its capabilities to cover search, time-series, graph data, and AI/ML inference, expanding its use cases far beyond simple caching.
From powering ecommerce personalisation and gaming leaderboards to enabling AI-driven recommendation engines and real-time analytics, Redis has established itself as the go-to database for speed, scale, and flexibility.

Installing Crossplane Providers 🚀

🎬 Act 1: The Foundation
Crossplane acts as a universal control panel for your infrastructure, but it needs “providers” to know how to talk to different clouds and services. Installing a provider is like plugging in a new extension cord so Crossplane can reach AWS, Azure, or wherever you want to provision resources.
```bash
kubectl crossplane install provider crossplane/provider-aws:v0.35.0
```What just happened?
Installed the AWS provider into your Crossplane instance
Registered new Kubernetes resource types (like S3Bucket, RDSInstance)
Made your cluster ready to orchestrate AWS resources
Like teaching your platform a new language!
🔑 Act 2: Credentials - Wiring in Permissions
Providers need secrets (credentials) to connect to your cloud accounts. Here’s how you house those credentials in Kubernetes:
kubectl create secret generic aws-creds \
--from-literal=key=YOUR_AWS_ACCESS_KEY \
--from-literal=secret=YOUR_AWS_SECRET_KEY \
-n crossplane-systemEssential breakdown:
Credentials are loaded into a Kubernetes Secret
Make sure the secret lives in the same namespace as Crossplane (often crossplane-system)
Minimal privilege: always restrict permissions in your cloud account for security
Like handing Crossplane a secure keyring to your cloud kingdom.
⚙️ Act 3: ProviderConfig - The Trust Bridge
Next, tell Crossplane where to find the credentials and how to use them. Enter the ProviderConfig, a YAML declaration that builds the bridge:
apiVersion: aws.crossplane.io/v1beta1
kind: ProviderConfig
metadata:
name: aws-default
spec:
credentials:
source: Secret
secretRef:
namespace: crossplane-system
name: aws-creds
key: keyWhat does this do?
Declares a named bridge (aws-default) for all AWS requests
Points the provider to your secret
Enables resource provisioning with your chosen credentials
Like telling your platform, “use these keys when opening AWS doors.”
✨ Act 4: Verification - Provider Health Check
Ensure your provider is healthy and connected:
kubectl get provider
kubectl describe provider <provider-name>Checklist:
STATUS should say “Installed” or “Healthy”
Errors will show if your secret is missing or has the wrong keys
If all is green, you're ready to deploy resources
This step is your systems check before the takeoff!
🚀 Final Act: Expanding the Universe
You can repeat these steps for other clouds (Azure, GCP), databases, or platforms by swapping the provider name and credentials. Each new provider expands what your platform team can offer, all through Kubernetes!
With Crossplane providers wired in, your platform delivers secure, scalable, and self-service infrastructure powered entirely by code.
🎯 Stay Inspired - Case Studies
Who: Spotify, the world’s leading audio streaming platform with over 270 million monthly active users and 124 million Premium subscribers across 79 markets, offering access to more than 50 million songs and podcasts.
What They Did:
To support its hyper-growth and innovation in music and podcast streaming, Spotify migrated its infrastructure from on-premises data centers to Google Cloud. The migration enabled rapid product development, deeper personalisation, and stronger artist-to-fan connections. Key initiatives included:
Migrated 1,200 online services, 20,000 daily job executions, and 100+ engineering teams to Google Cloud
Adopted BigQuery, Pub/Sub, and Dataflow for real-time and historical data analysis
Built dashboards for artists to understand fan engagement and plan tours or releases
Enhanced personalisation features with machine learning while safeguarding customer privacy
Scaled global R&D operations across multiple continents with a collaboration-driven engineering culture
Tech Stack & Tools Used: Google Cloud Platform, BigQuery, Pub/Sub, Dataflow, Google Cloud networking and compute services, ML and analytics tools
Why It Matters:
✅ Massive scale with billions of daily streams supported by resilient, global cloud infrastructure
✅ Faster innovation with automated services freeing engineers to focus on product features and customer experience
✅ Personalisation at scale with advanced ML and analytics powering tailored listening and discovery experiences
✅ Artist empowerment through dashboards that help creators connect with fans and make smarter touring and release decisions
✅ Sustained growth with 29 percent Premium subscriber growth in a single year driven by new markets and innovative products
👀 In Case You Missed It…
Agentgateway Donated to CNCF to Accelerate Secure AI Agent Adoption Agentgateway, an AI-native proxy created by Solo.io, has been donated to the Linux Foundation to accelerate secure and governed AI agent adoption, with backing from major contributors like AWS, Microsoft, and Red Hat.
Outdated Python Versions Cost Companies Millions
Running older versions goes beyond being a technical debt problem; it's a financial haemorrhage bleeding out organisations’ cloud bills.
Microsoft: AI Agents Are Winning the DevOps Relay Race
Microsoft says AI agents are transforming software development from a fragmented relay race into unified workflows.
Till next time,