Xin Y.

We’re two days into AWS re:Invent 🎄. If you’re running Lambda, OpenSearch, EKS, or any managed databases, this post is for you 👀 ⭐ 𝗟𝗮𝗺𝗯𝗱𝗮 𝗠𝗮𝗻𝗮𝗴𝗲𝗱 𝗜𝗻𝘀𝘁𝗮𝗻𝗰𝗲𝘀 If you’re spending thousands on Lambda and Fargate isn't for you, this might be. You can now run Lambda functions on EC2 instances with whatever hardware you fancy - GPUs, niche CPUs, more memory than sense - and past the old 15-minute limit. Instances are eligible for savings plan and there's no per-ms execution cost but that 15% management fee is going to make this a PITA to calculate. ⭐ 𝗟𝗮𝗺𝗯𝗱𝗮 𝗗𝘂𝗿𝗮𝗯𝗹𝗲 𝗙𝘂𝗻𝗰𝘁𝗶𝗼𝗻𝘀 Step Functions, but in code and without the feeling you’re wiring together a Victorian telegraph system. Flows can pause, resume, and generally behave like modern workflows. Pricing is a bit of a puzzle, but the model is genuinely impressive - closer to Cloudflare Workflows or Temporal than the AWS of old. ⭐ 𝗢𝗽𝗲𝗻𝗦𝗲𝗮𝗿𝗰𝗵 + 𝗚𝗣𝗨𝘀 For anyone rebuilding large vector indexes: GPUs are now an option. Running them 24/7 will cost more than your London rent, but as part of a batch pipeline they can give you 4× faster builds and a cleaner conscience. ⭐ 𝗘𝗞𝗦 𝗨𝗽𝗱𝗮𝘁𝗲𝘀 Not a big Kubernetes user myself, but friends who are neck-deep and running ArgoCD, ACK, or KRO say this release might actually reduce operational pain. I’ll believe it when I see a Kubernetes upgrade that doesn’t require lighting incense. ⭐ 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲 𝗦𝗮𝘃𝗶𝗻𝗴𝘀 𝗣𝗹𝗮𝗻𝘀 At long last. Savings plans for databases (up to 35% for 1-year commitments). Coverage includes Aurora, RDS, DynamoDB, ElastiCache, DocumentDB, Neptune, Keyspaces, Timestream, and DMS. They even apply to serverless and on-demand, which feels uncharacteristically generous. I’m sure there’s a catch somewhere, but let’s enjoy the moment. Loads of other announcements here: https://lnkd.in/eAZAPD_u

4

Are AI agents entering their "Kubernetes" era? Databricks open-sourced Omnigent—a "meta-harness" designed to completely change how we orchestrate AI agents. If you’ve been building or coding with agents lately, you know the struggle. Today’s agent harnesses (like Claude Code, Codex, or Pi) are amazing, but they operate in silos. You find yourself copy-pasting code between tools, juggling different environments, and struggling to manage overall cost and security. Databricks (led by Matei Zaharia and Kasey Uhlenhuth) built Omnigent to solve this. It introduces a brand new abstraction layer that sits above individual agent harnesses, turning them into interoperable, team-oriented workers. Key features that make this a game-changer: 🔹 Composition: Mix, match, and swap between Claude Code, Codex, Pi, and custom agents with single-line changes—no massive code rewrites. 🔹 Control (No More Prompt Guardrails): Enforce stateful, contextual policies like spend budgets and workspace permissions directly at the meta-harness layer, not through finicky prompt engineering. 🔹 Live Collaboration: Share active agent sessions via URL. Teammates can watch agents work in real-time, co-drive execution on your machine, or fork conversations. 🔹 Multi-Device Sync: Seamlessly jump from your terminal to web, desktop, and even your phone, with all files and chat history completely in sync. The best part? It’s officially open-source (Apache 2.0) and ready for alpha testing. Check out the full breakdown through link in comment👇 #AIAgents #GenerativeAI #OpenSource #Databricks #SoftwareEngineering #TechInnovation

3

1 Comment

AWS is promoting rust on lambda to GA https://lnkd.in/g-YPXh_f , but it's too little too late - you can easily run rust unikernels for far less $$, way faster performance, less cloud lock-in and a lot more security.

22

Amazon MSK (Managed Streaming for Apache Kafka) is the AWS service who managed kafka broker Before to start with Amazon MSK read this article. This article can help begenners and senior professional kafka architects on AWS.

2

The best advice I got on my promotions as a software engineer was from a Principal Architect at Amazon after he told me his story of getting promoted to Principal, and this has stayed with me my whole career. He said, very simply: > Your architecture can be a love letter to your technical brilliance, > but if it does not move customer and company goals, it will not move your level. That hurt a little. Because I recognised myself in it. A few years ago I wanted to design something with every shiny thing in the book: API Gateway, Lambda, DynamoDB, event bus, the works. On paper it was beautiful. In reality it was expensive, slow to ship, and did not match the skills of the team that would maintain it. So we shipped a hybrid architecture instead. Fewer moving parts, more reuse of existing services, cheaper to run, easier to hand over. Over time, here is how my thinking about promotions changed. 1. Start from the business goal, not the tech stack Before talking about Kafka vs SQS or MySQL vs DynamoDB, answer this first: What money, risk, or customer pain will this remove? Promotions track that, not the number of services in your diagram. 2. Ship finished products, not clever demos Leaders remember the feature that went live, stayed stable, and earned or saved real money. Drive the entire lifecycle yourself: design, build, launch, oncall, cleanup, docs. 3. Make the team faster, not just yourself Tools, scripts, templates, better CI, better dashboards. If people quietly say “things move faster when you are on this” your manager already has half a promotion case. 4. Do your share of the unsexy work Oncall, migrations, bugs, incident reviews. Senior folks who avoid this slow the entire org. Senior folks who lean into it become trusted very quickly. 5. Grow people, not just systems Mentor juniors, run design reviews, write clear docs, set standards. When your manager sees that you are already acting like the next level, formal promotion becomes a catch up, not a favour. 6. Be proactive about your promotion, not obsessed with it Have periodic one to ones. Talk about where you want to go and ask what evidence your manager would need to support that. Remember your manager has ten other people to manage. It is your responsibility to make your impact visible. Do not be desperate. Understand the timelines and work within them. As you get close to the boundary, collect the missing data points: impact metrics, incident write ups, design docs, peer feedback. Make the story easy to tell. Most importantly, work for the business, not for promotion-driven development. If your roadmap only exists to tick boxes on a promotion rubric, people can feel it. If your roadmap clearly creates value, saves cost, or reduces risk, promotion becomes the natural side effect. So instead of asking “What new tech can I use this year?” Ask, “Where can I create obvious business impact and leave the system and the people around me better than I found them?”

182

15 Comments