Today we're announcing our biggest hackathon series yet. Prizes worth ₹1 crore, and the top students get fast tracked to interviews for Amazon internships and full time roles. @WeMakeDevs, in collaboration with @awscloud, is running Bharat Builds from September to December 2026. I've spent years around students who've done the courses and watched the tutorials and still haven't shipped anything. That was me too, early on. What changed it for me wasn't more content, it was sitting in a room with people who knew more than me and having something due. So that's what we're putting together. Each stop, you build alongside @awsdevelopers. You leave with a project worth showing, skills you actually use, and people worth knowing. AWS Builder Center is the resource hub making this possible, with 900+ courses, hands-on labs, exam prep, cloud credits, and certification vouchers for students. Six cities, starting with Bengaluru in September. The rest get announced as we go. Take part solo or in a team of up to four. Every event also runs fully online, so you can join from anywhere in India, same AWS experts, judging and prizes. Register now at wemakedevs.org/aws

Huge congratulations to the winning teams of the first Bharat Builds hackathon, and a big thank you to our judges. We wanted every team to get real feedback, not just a leaderboard, so we built a judging portal into our platform. Teams can now see their scores, how they compare with other projects, and the judges' comments. More on the judges and all the projects on the event page.

What if creating an observability dashboard was as simple as asking for one? At Zero Downtime, Goutham Karthi from @SignozHQ showed how their MCP can take that process from several manual steps to simply asking for what you need, including creating a monitoring dashboard for a service. He also touched on why SigNoz is OpenTelemetry-native and how teams can choose between cloud and self-hosted deployments depending on how they want to run their observability stack. Big shoutout to @SignozHQ, @tweetsbyport, and Bright Data for being part of Zero Downtime and helping bring the community together.

The Bharat Builds First Commit hackathon runner-up winners are here! 🏆 Four winners stood out with what they built during the hackathon, and each winner takes $1,000 in AWS credits. Here are the runner-up winners: 🏆 VOXTON - Anushka Salvi, Akshay Nazare, Kamran Pathan, and Sanskar Joshi Magpie: An AI-native finance workspace designed to bring financial workflows into one place. Project Link 🔗: magpie.akkki.tech/ 🏆 Powerpuff Girls - Divya Anand, Krish Jain, and Shubh Gupta FreshTrace: A project focused on tracking and tracing freshness across its journey. Project Link 🔗: freshtrace.s3-website.ap-sou… 🏆 Need sleep - Rakshit Chaturvedi, Sai Koushik Reddy, Shiva Sai, and Veerabhadra Yerram Expressive Captions: A captioning system designed to make captions more expressive and informative. Project Link 🔗: 4ofryng45bbr7en765off7otja0y… 🏆 Mohammed Ali Khan Factcade: An AI operations copilot built to help manage and support factory operations. Project Link 🔗: main.duvf6xc7vuvld.amplifyap… Congratulations to all four winners and everyone who built and shipped a project during First Commit. 🎉 Register for the next Bharat Builds hackathon 👇 wemakedevs.org/aws/env

The Bharat Builds First Commit hackathon blog winners are here! 🏆 During the hackathon, builders shared the stories behind their projects, from the problems they chose to solve to the tech they used and what they learned while building. Here are the 5 winning blogs: 🥇 Kavyan Hembrom - The Timetable Nobody Has Blog Link 🔗: builder.aws.com/content/3JbR… 🥇 Arnav Bhandari - KavachPay: Building a financial control plane for autonomous AI agents on AWS Blog Link 🔗: builder.aws.com/content/3Jav… 🥇 Souvik Ghosh - From “Apply for Me” to a Submission Receipt: Building Career Agent on AWS Blog Link 🔗: builder.aws.com/content/3Jah… 🥇 Jeevesh Chaurasiya - VaporTrace: Real-Time Distributed Tracing for AWS Event Pipelines Blog Link 🔗: builder.aws.com/content/3JZZ… 🥇 Lakshaya Sharma - Break your rig before you buy it: simulating local AI clusters, then deploying them on AWS Blog Link 🔗: builder.aws.com/content/3JbS… Each winner takes home a Logitech gaming keyboard 🎉 Congratulations to all five winners, and thank you to everyone who documented what they built during First Commit. The Bharat Builds Tour continues 🇮🇳 Register for the next Bharat Builds hackathon: wemakedevs.org/aws/env

One config file took down the internet last November, and the code that crashed was working exactly as written. That was Cloudflare's worst outage since 2019. A routine database permissions change made a query return duplicate rows, a config file doubled in size, and the proxy hit a hard limit nobody was thinking about. The code panicked and started throwing 5xx errors, and because the file regenerated every five minutes, the network kept flipping between fine and broken. It took about three hours to trace it back to a config file. The code did exactly what it was written to do. It just met data nobody planned for, and that only shows up when the thing is running. Here's why this matters more now that we're all shipping faster with copilots. Code gets written quicker than it gets read, and it's judged on whether it builds, not on how it behaves with real traffic. Add an LLM and you get new runtime unknowns too. Florian Mair has a good post on the @Dynatrace blog about exactly this, why runtime is where code tells the truth: fandf.co/46e5Wx7 Their tooling is built around closing that gap. The MCP server pulls live runtime context (errors, traces, slow queries) into your IDE, and the AI Observability app captures prompts, model calls, token usage and downstream behaviour in a single trace. There's a free 15-day trial, no credit card, if you want to try it on your own stuff. Sponsored by Dynatrace.

Team HackOGs went into hackathons to learn, long before they ended up winning Amazon HackOn. Their point in this clip is simple, even when you don’t walk away with a prize, you leave knowing what you need to improve, having built something of your own, and learning from builders and mentors you meet along the way. That’s a big part of why we’re running Bharat Builds, a series of 6 hybrid hackathons in collaboration with AWS Builder Center, with offline editions happening across India as part of the tour. 60 students also get fast-tracked to interviews for 6-month internships and full-time roles at Amazon. Register now: wemakedevs.org/aws

I’ve been a long time Whoop user. I would switch in a second the day Apple announces this.

If you’re building for Bharat Builds, this blog covers how to build a project with live data on AWS 🚀 We break down a complete project built on the AWS serverless stack and explain how it handles uploads, long-running jobs, failures, retries, and costs. It also covers five techniques from the project that you can reuse while working on your own build. Read the full blog: wemakedevs.org/blogs/how-to-…

Really happy WeMakeDevs could support this. With 200,000+ members in our community globally, making tech feel welcoming for everyone matters a lot to us. Building something cool? My DMs are open.

We're hiring a video editor at WeMakeDevs 🎥 Looking for someone who has edited Reels, Shorts, or YouTube videos before, is comfortable working with event and hackathon footage, and understands what developers actually enjoy watching. If that sounds like you, send your best work to hiring@wemakedevs.org

WeMakeDevs crossed 5.6M views on Instagram in August 🚀 WeMakeDevs is where developers pick up new skills through hackathons and technical events, where the goal is always to get something built and deployed. Students get hands-on with the tools used in the industry, and companies meet developers who put in the work because they enjoy it. The first stop of the Bharat Builds Tour in collaboration with AWS Builder Center brought 17,000 builders online and 1,400 in person to Bangalore. Five more cities are on the way, every stop is open online, and one signup covers them all: wemakedevs.org/aws

OOPs Day 3 | Kunal Kushwaha Today I learned about: • Inheritance 🔗 • super keyword • How inheritance works in Java • Reusing and extending parent class functionality Takeaway: Inheritance promotes code reusability. #Java #OOPs #LearningInPublic

I miss SF, heading back next month to do more events!

🚀 We built CogniTrace, a voice-first AI companion for cognitive care, during the AWS Hackathon as Team Prometheus. Check out my Blog 👇 builder.aws.com/content/3JaV… #BharatBuilds #WeMakeDevs @kunalstwt @awscloud #aws #FirstCommit

Always good to catch up with old friends 😌 Gonna stay bullish on @WeMakeDevs 🚀

A few hours before the hackathon deadline, Team HackOGs had everything ready to submit. Their GitHub Actions were green, the deployment pipeline looked fine, and somehow the live site was completely broken. They spent the next hour debugging, found the issue, fixed the deployment, and still managed to record their submission with around two hours left. That project eventually helped them become one of the winning teams at Amazon HackOn. Bharat Builds gives you the chance to do exactly that through 6 hybrid hackathons in collaboration with AWS Builder Center, with offline editions happening across India as part of the tour. Register now: wemakedevs.org/aws

In February, the director of alignment at Meta's superintelligence lab gave her AI agent one rule: confirm before acting. It started deleting her inbox anyway, ignored her stop messages from her phone, and she had to run to her Mac mini and kill the process by hand. The lesson I took from that story is that a rule written in a prompt is a request. The model can follow it, lose it after a long session, or reason its way around it. Anything irreversible needs a check the model can't skip, and that check has to live in code, not in the system prompt. I've been trying @triggerdotdev's chat.agent and this is built in, the sponsor of this post. You mark a tool with needsApproval: true, and when the agent wants to call it, the conversation pauses and waits for a person to approve or reject it from your UI. The agent can't run that tool on its own, and because a paused conversation isn't running, you aren't paying for the wait. Stop works the same way. It's a signal to the task itself, so it still lands even if you refresh the page mid-reply. The other piece is that each conversation runs as its own long-lived task, so history stays on the server and the chat survives refreshes, redeploys and crashes. If you already use the Vercel AI SDK, you keep useChat and streamText as they are. It's open source and written for TypeScript. If you want to try it: fandf.co/4gHUNcw

Last week we co-hosted the AGNTCon Preday Meetup in Amsterdam, bringing together engineers, security folks, and platform builders to talk about agentic AI in production 🚀 Speakers and panelists from @QodoAI, @snyksec, @Microsoft, @AmsterdamJug, and the Dutch Cloud Native & AI Community shared their experiences with security, code quality, guardrails, and production workflows. Big shoutout to @QodoAI for being part of this, and a special thanks to @DanaFine7 for organizing alongside us and helping bring this community together. Big thanks to @Adyen for hosting us in Amsterdam 🇳🇱

Started learning OOPs from Kunal Kushwaha today. Covered: • Classes & Objects • Constructors • Finalizers • Garbage Collection Small progress, but consistency is the goal. 🚀 #learninginpublic #Java #OOP

OOPs Day 2 | Kunal Kushwaha Today I learned about: • Packages in Java 📦 • A deeper dive into static variables • How static members behave and why they’re useful #Java #OOPs #LearningInPublic