Pinned Tweet
One comment I get all the time going from event to another "How do we keep up in this crazy world of AI, Agents and our jobs changing daily".
Going away for vacation for 1 month and coming back to a full schedule of back to back events I also started questioning myself about ood enough anymore!
I took time to reflect on it. Checked couple of things, spoke to a few colleagues I trust and came to the following conclusion "It doesn't matter"!
What matters is focus on the things you need to need when you need to know them. Frameworks that were cool 2 months ago are gone, no one talks about them anymore, LLMs that were a hype one month ago are not available for pennies on the dollar for millions of tokens.
The only constant is change. This are moving fast and it will take time for any one mode, framework, agent, harness, cli to stick around
Embrace the chaos 🙌
PS: Photo is irrelevant to this post. It's from climb to Ararat. I just wanted people to see this post 😂
Going away for vacation for 1 month and coming back to a full schedule of back to back events I also started questioning myself about ood enough anymore!
I took time to reflect on it. Checked couple of things, spoke to a few colleagues I trust and came to the following conclusion "It doesn't matter"!
What matters is focus on the things you need to need when you need to know them. Frameworks that were cool 2 months ago are gone, no one talks about them anymore, LLMs that were a hype one month ago are not available for pennies on the dollar for millions of tokens.
The only constant is change. This are moving fast and it will take time for any one mode, framework, agent, harness, cli to stick around
Embrace the chaos 🙌
PS: Photo is irrelevant to this post. It's from climb to Ararat. I just wanted people to see this post 😂
0
0
13
499
Your idle GKE workloads are costing you money! This means you have to manage complex add-ons to make fix it? Well we have a fix! 🚀
GKE 1.37 just dropped native Scale-to-Zero capabilities, and it is a step forward for cloud-native elasticity. No more paying for always-on infrastructure just to keep event-driven workers or dev environments ready.
Here is why I’m so excited about this one:
📉 Zero Replicas, Zero Waste: Standard HPA can now hit minReplicas: 0. When the queue is empty, the pods are gone.
🛠️ Goodbye Add-on Overhead: We moved the logic directly into the GKE control plane. You don't need KEDA or heavy sidecar operators anymore. Just a simple AutoscalingMetric CRD. Less complexity, way less YAML!
⚡ No More Cold Starts: By combining this with GKE Capacity Buffers, your pods have pooled, warm nodes waiting for them. You get instant scale-up when demand returns, completely slashing that 60-90 second node provisioning latency.
True elasticity is about decoupling your infrastructure costs from workload readiness. The days of paying for idle resources are numbered! 🚢
Check out the full blog post to see how to configure your first native scale-to-zero workload today. 👇
cloud.google.com/blog/products/…
#Kubernetes #GKE #GoogleCloud #CloudNative #PlatformEngineering
GKE 1.37 just dropped native Scale-to-Zero capabilities, and it is a step forward for cloud-native elasticity. No more paying for always-on infrastructure just to keep event-driven workers or dev environments ready.
Here is why I’m so excited about this one:
📉 Zero Replicas, Zero Waste: Standard HPA can now hit minReplicas: 0. When the queue is empty, the pods are gone.
🛠️ Goodbye Add-on Overhead: We moved the logic directly into the GKE control plane. You don't need KEDA or heavy sidecar operators anymore. Just a simple AutoscalingMetric CRD. Less complexity, way less YAML!
⚡ No More Cold Starts: By combining this with GKE Capacity Buffers, your pods have pooled, warm nodes waiting for them. You get instant scale-up when demand returns, completely slashing that 60-90 second node provisioning latency.
True elasticity is about decoupling your infrastructure costs from workload readiness. The days of paying for idle resources are numbered! 🚢
Check out the full blog post to see how to configure your first native scale-to-zero workload today. 👇
cloud.google.com/blog/products/…
#Kubernetes #GKE #GoogleCloud #CloudNative #PlatformEngineering
0
0
0
205
Abdel SGHIOUAR retweeted
And we crushed it!!! 🇲🇦
Just delivered the opening keynote "The Era of Doers" at @DevoxxMA, and the feedback has been amazing.
Truly honored to be here. Thank you @boredabdel and @badrelhouari for having me!
Grazie, merci, shukran 🙏
#DevoxxMA
Just delivered the opening keynote "The Era of Doers" at @DevoxxMA, and the feedback has been amazing.
Truly honored to be here. Thank you @boredabdel and @badrelhouari for having me!
Grazie, merci, shukran 🙏
#DevoxxMA
1
2
59
2,474
Our first keynote speaker @FrancescoCiull4 on stage @DevoxxMA 🇲🇦
Talking about how we are in an era of Doers instead of over thinkers with AI 🤖
#DevoxxMA
Talking about how we are in an era of Doers instead of over thinkers with AI 🤖
#DevoxxMA
0
1
10
691
As KubeCon NA co-chair, I can't wait to see everyone in Salt Lake City in less than two weeks!
Let’s swap production stories, talk cloud native, and maybe discuss a few "on air" Kubernetes puzzles, peacefully, of course. 🎙️
Get my 25% discount code for KubeCon passes in the CloudNativeCon #KubernetesPodcast
Let’s swap production stories, talk cloud native, and maybe discuss a few "on air" Kubernetes puzzles, peacefully, of course. 🎙️
Get my 25% discount code for KubeCon passes in the CloudNativeCon #KubernetesPodcast
2
1
7
903
Use code KCNA26AMB25 at registration to save 25% on your KubeCon-only pass. Let me know if you are coming so we can connect!
Register here: events.linuxfoundation.org/kubecon-cloudn…
Register here: events.linuxfoundation.org/kubecon-cloudn…
0
0
0
85
🚀 Antigravity Unlocked: Meet the Team Behind the Product in Stockholm! 🚀
Want to know what the future of agentic development looks like? Get an inside look directly from the source. On Monday, November 16, we welcome Damith Karunaratne, Group Product Manager for Google's AI ckholm!
This is an exclusive opportunity to dive into Google Antigravity, explore the tools shaping AI-driven engineering, and discuss the future of development with the team building it.
Don't miss out on this deep dive and the chance to connect with fellow innovators.
📅 Monday, November 16, 2026
🕒 15:00–18:00 (Networking until 19:00)
🏢 Google Cloud, Kungsbron 2, Stockholm
🔗 Register here: rsvp.withgoogle.com/events/antigra…
#GoogleAntigravity #ArtificialIntelligence #AIagents #GoogleDeveloper #StockholmTech #TechEvent
Want to know what the future of agentic development looks like? Get an inside look directly from the source. On Monday, November 16, we welcome Damith Karunaratne, Group Product Manager for Google's AI ckholm!
This is an exclusive opportunity to dive into Google Antigravity, explore the tools shaping AI-driven engineering, and discuss the future of development with the team building it.
Don't miss out on this deep dive and the chance to connect with fellow innovators.
📅 Monday, November 16, 2026
🕒 15:00–18:00 (Networking until 19:00)
🏢 Google Cloud, Kungsbron 2, Stockholm
🔗 Register here: rsvp.withgoogle.com/events/antigra…
#GoogleAntigravity #ArtificialIntelligence #AIagents #GoogleDeveloper #StockholmTech #TechEvent
1
0
7
515
Abdel SGHIOUAR retweeted
Next up on the #DevoxxMA 2026 stage: @fniephaus 🎤
Get ready for his talk: “GraalVM meets WebAssembly” 💻⚡
🎟️ Get Your Ticket Now: devoxx.ma
#DevoxxMA
Get ready for his talk: “GraalVM meets WebAssembly” 💻⚡
🎟️ Get Your Ticket Now: devoxx.ma
#DevoxxMA
0
1
6
776
Abdel SGHIOUAR retweeted
@mstdayoub is joining us on the #DevoxxMA 2026 stage! 🎤🚀
Get ready for his talk: “You Can’t Debug Your Health with More Code”
🎟️ Get Your Ticket Now: devoxx.ma
#DevoxxMA
Get ready for his talk: “You Can’t Debug Your Health with More Code”
🎟️ Get Your Ticket Now: devoxx.ma
#DevoxxMA
0
2
8
913
Abdel SGHIOUAR retweeted
Who will I be seeing next week? Because that’s where you’ll find me chatting about #eXtremeProgramming (and, of course) a bit about #AI 😉
And –needless to say– really looking forward to getting to know the local community 🙂🧑💻👩🏽💻👨🏿💻
#DevoxxMA
And –needless to say– really looking forward to getting to know the local community 🙂🧑💻👩🏽💻👨🏿💻
#DevoxxMA
🚀 7 days until Devoxx Morocco 2026!
Two days of tech, learning, and connections with the developer community.
📅 November 4–5
📍 Casablanca
🎟️ Get your pass: devoxx.ma
#DevoxxMA #DevoxxMorocco
1
2
7
375
Abdel SGHIOUAR retweeted
🚀 7 days until Devoxx Morocco 2026!
Two days of tech, learning, and connections with the developer community.
📅 November 4–5
📍 Casablanca
🎟️ Get your pass: devoxx.ma
#DevoxxMA #DevoxxMorocco
Two days of tech, learning, and connections with the developer community.
📅 November 4–5
📍 Casablanca
🎟️ Get your pass: devoxx.ma
#DevoxxMA #DevoxxMorocco
0
2
13
1,081
Abdel SGHIOUAR retweeted
Your developer platform isn't just for humans anymore. How can AI agents access applications and infrastructure while respecting guardrails?In this video, @boredabdel explores the OpenChoreo MCP Server.
Watch the full video:
youtube.com/watch?v=LbHmfT…
Watch the full video:
youtube.com/watch?v=LbHmfT…
0
2
2
73
Abdel SGHIOUAR retweeted
After an incredible 7-day break from traveling...
the Italian boy is back at the airport!
Today: 🇮🇹 Rome → Casablanca 🇲🇦 with @Giulia_Trinidad
Wednesday: opening keynote at @DevoxxMA.
Let's CRUSH it!
the Italian boy is back at the airport!
Today: 🇮🇹 Rome → Casablanca 🇲🇦 with @Giulia_Trinidad
Wednesday: opening keynote at @DevoxxMA.
Let's CRUSH it!
7
1
53
2,720
Tired of slow Pod cold starts or over-provisioning CPU just to get your apps to initialize quickly?
My new post about GKE CPU startup boost a feature that solves this problem.
Instead of paying for permanent CPU overhead, startup boost dynamically increases a container’s CPU h and automatically scales it back down to your baseline once the app is ready.
It does this live, with zero pod restarts, using Kubernetes In-place Pod Resize (IPPR) via GKE's Vertical Pod Autoscaler. With this you get:
✅ Up to 2x faster cold starts for CPU-intensive apps (Java, Node.js, Python).
✅ Lower costs stop paying for steady-state over-provisioning.
✅ Flexible policies using simple multipliers or container-specific rules
Stop sizing for startup and paying for it forever. Read the post to see how to enable it on GKE Standard and Autopilot: 🔗 cloud.google.com/blog/products/…
#GKE #Kubernetes #GoogleCloud #FinOps
My new post about GKE CPU startup boost a feature that solves this problem.
Instead of paying for permanent CPU overhead, startup boost dynamically increases a container’s CPU h and automatically scales it back down to your baseline once the app is ready.
It does this live, with zero pod restarts, using Kubernetes In-place Pod Resize (IPPR) via GKE's Vertical Pod Autoscaler. With this you get:
✅ Up to 2x faster cold starts for CPU-intensive apps (Java, Node.js, Python).
✅ Lower costs stop paying for steady-state over-provisioning.
✅ Flexible policies using simple multipliers or container-specific rules
Stop sizing for startup and paying for it forever. Read the post to see how to enable it on GKE Standard and Autopilot: 🔗 cloud.google.com/blog/products/…
#GKE #Kubernetes #GoogleCloud #FinOps
2
0
2
292
