First participation and first win! Didn’t expect it, but hard work (and a bit of luck since I started 10 days late) pays off. Not the main track, but a win is a win. More to come 🚀
And because the submissions were that good…
Five extra projects earned Noteworthy Mentions 🎉
1. Chain Pilot by @Chain_Oracle
2. BlockIQ by @shahmusk12
3. ReflectIQ by @riteshkrkarn & @Khushim1109
4. KeeperDCA by @localhost_ayush
5. Mendel AI by @alienworl1
Each project will be awarded $200 for their noteworthy contributions 👏👏
After you learn to code, the most important skill is product thinking.
Coding is table stakes now. Tokens keep getting cheaper, so that's the easy part.
The real flex is knowing what to build. Building the right thing is 10x harder than building the thing right.
@RahulKrishnaa28 If you're a web dev who ships, you already know the basics: spacing, layout, and breakpoints. Prompting that knowledge > learning Figma. One skill compounds everywhere; the other doesn't.
If you're a dev who can ship but can't design: learning to prompt AI well is higher leverage than learning Figma. Prompting compounds everywhere. Figma only helps with design.
Researching auth for production taught me this: blocking a user's account doesn't kill their active session.
They can't log in again, but they're still authenticated until the token expires.
Solution: short-lived JWTs or maintain revoked tokens in Redis.
First week lesson: gitflow. Not complicated, just organized git. But that's the point.
Production code needs systems, not just skills. Follow conventions when you're the only dev so the code is ready when you're not.
Tutorials teach commands. Teams need discipline.
Tried building with just HTML, CSS, and vanilla JS today and realized I’d forgotten some very basic things. Funny how learning new stacks can make you overlook how important the fundamentals really are.
Tried Antigravity today and it’s one hell of a tool. From setup to getting a good project running, took just a few hours with some detailed prompts and a few iterations. Everything top AI models and features is free for now. Use it while it lasts. Google really cooked this time.
Finally got the Projects update in ChatGPT, and it fixes my biggest frustration with AI not remembering context. No more rewriting prompts or hitting limits for long tasks. Now exploring it deeply and iterating to get the best out of it.
Go middlewares looked scary at first with all those layered function wraps. But once the flow clicked, it felt so intentional and simple. One of those things that looks complex until it isn’t.
After winning an agentic AI hackathon and completing the Kaggle course, the shift finally clicked for me. We’ve moved from AI that only responds to AI that actually gets things done. Definitely piqued my interest to build more in this space.
Kaggle's 5-day AI Agents Intensive course really lives up to its name: one new research paper, lecture and code labs every day. But honestly, that's a good push for me. I've been thinking of getting into regular paper reading for a while.
After learning about threads, I realized balance works the same way: it's not doing everything at once; it's just switching fast enough to make it look like you are.
Most people think:
GET → read
POST → create
PUT → update
DELETE → delete
But that's just convention.
HTTP methods only define what you intend to do, not what your backend actually does.