My Experience with September 2025
A recap of my exciting tech journey in September 2025
What an incredible September it’s been! This past month, I’ve been diving deep into some cutting-edge tech and exciting new projects. Here’s a quick recap of my highlights:
Cloud-Native & DevOps Adventures:
- Spearheaded the Email support with AWS SES integrations and came to know about DMARC, DKIM, and SPF configurations.
- Explored advanced Terraform modules for Cloudflare, focusing on email routing and Google Groups integration.
- Delved into the intricacies of
helm upgradefor existing releases usingksctl/ksctlgo package.
Performance & Sustainability Focus:
- Gained extensive knowledge on RAPL and non-RAPL estimators for energy consumption, tackling challenges in cloud VM environments.
- Pioneered platform-level addon support, with a successful POC for per-process/container energy consumption metrics.
- Researched best practices for AI Agent tool creation from Anthropic and dived into Nvidia GPU operator blog posts.
- Explored the nuances of grid carbon intensity (average vs. marginal) and the core, uncore, and package zones for RAPL.
- Worked extensively with Kepler, Prometheus, and green metrics, including OCI-based Helm deployments.
System Design & Core Infrastructure:
- Mastered raw HTTP calling of the K8s API server using
rest.Configandrest.Client. - Explored
devicemapperand thin loop devices, understanding their role incontainerdfor image storage. - Learned about
firecrackerfor secure application delivery and the fascinatinggossip protocol. - Deepened my understanding of
fluxcdmetalabels andpromQLfiltering.
Product Development & Community:
- Contributed to new UI/UX designs and tackled complex Stripe checkout and quota system edge cases.
- Explored state machines for real-world project applications.
It’s been a month of intense learning, building, and contributing! Stay tuned for more updates on these exciting developments.
#TechRecap #SeptemberHighlights #Kubernetes #DevOps #CloudNative #GreenTech #SystemDesign #ksctl #AI #LearningAndGrowing
Other posts
I Switched My Local Copilot From llama.cpp to vLLM — 2× VRAM for 50% Faster Completions
↗I swapped llama.cpp for vLLM on my local code completion setup. 49% faster generation, 36% lower latency — but 2× the VRAM. Here's what I gained, what it cost, and the gritty install details.
I Built a Copilot Clone in Neovim With a 1.5B Model on a Laptop GPU
↗Using Qwen2.5-Coder-1.5B with llama.cpp and Minuet AI to get Fill-in-the-Middle code completions in Neovim — no subscription, no telemetry, running on a laptop GPU.
Building My First SaaS: Kubmin in the AI Agent Era
↗What building Kubmin taught me about SaaS system design, observability, deployment, queues, auth, and using AI agents without losing ownership.