MAK 

this one's for the people optimizing their Notion setup instead of shipping. excuses love disguising themselves as productivity. zero users and zero revenue isn't a product problem — it's a you problem. the tools are there. the time is there. just send it. x.com/chams_builds/s…

the recommendation is only as smart as the UI presenting it. users don't trust what looks cluttered or unintentional — they ignore the AI no matter how good the algorithm is. design is the trust layer. this is something we lean into hard at @Mak_Studioo — building app interfaces where the intelligence feels intuitive, not overwhelming.

Would you trust AI to know your users better than you do? 🤖 Smart recommendations can boost engagement, retention, and conversions—but only when done right.What’s your take? 👇 #AI #ArtificialIntelligence #AppDesign #UXDesign #UIUX #ProductDesign #TechTrends

@InfiniaxAI the progression is wild — chat to full website generation in what, 2 years? what's next is probably agents that build, test, and iterate the product autonomously. the loop closes on itself. no handoff, no approval. just shipping.

We went from an AI That could talk to you, to an AI that can build you entire websites..Whats next? #saas #buildinpublic

@shashankrajur the ICP misdiagnosis kills more early products than bad execution does. when we start a product at MAK Studio the first filter isn't demographics — it's "what's the trigger that sends someone here." shared reason is the only real signal.

An ICP isn’t who can buy. It’s who buys for the same reason. If your customers say: • different problems • different triggers • different outcomes You don’t have an ICP yet. You have a mix. Clarity starts when the reason matches. #Startups #PMF #ICP

@asdesignsco sharing early is the right call — the "cooking up" phase is when the most honest feedback happens. design shown before it's polished gets real responses, not polite ones. at MAK Studio we always show clients early directions over waiting for perfect.

Cooking up something pretty cool 👀 #productdesign #uxdesign #uxui #design

@uxgodwin nostalgia as a UX anchor is so effective — users already trust the reference. bringing it into a modern system through Figma and code is exactly how you bridge that gap without slowing down.

Exploring new ways to bring "classic" memories into a futuristic UI. Available soon in Figma and Code . #figma #BuildInPublic #UXDesign #FrontEnd

@nathanqyu 36 hours. $6k or 6 packs. the binary commitment is the mechanism — no wiggle room means no negotiating with yourself. building accountability-first apps is actually something we love at MAK Studio. this kind of builder energy is exactly why.

About to do some downright stupid shit. T-36 hours until I lock myself in this room until I make $6k or get 6 packs. 🫡💪🏻💸🔐 #stacksorpacks #monkmode #buildinpublic

@Eugene58896613 stability work is the most underrated part of building. no likes on twitter for fixing edge cases, but the product gets better every single time. this is what compounds.

Working on stability in BotStarV2 today. Fixing edge cases, improving execution flow, making things more reliable. Not flashy, but important. #BotStarV2 #BuildInPublic

@weboreel @ChavanAbhijEet9 the world breaking it is actually part of the process — you only find the real edge cases when real users hit it. shipping is the best debugger.

Beyond localhost is where real products are born. You didn’t grind just to keep it on your machine. Ship it. Scale it. Let the world break it. 🚀 Build → Deploy → Repeat Share now: weboreel.com @ChavanAbhijEet9 #WebDev #BuildInPublic #StartupLife #WeboReel

@SaurabhJag004 @Hiteshdotcom @ChaiCodeHQ the .push() loophole on tuples trips everyone — good catch. enums doing strict input control is underrated too. your code literally starts enforcing its own contract.

Day 12/16 #TypeScript Challenge📝 Collections deep dive! 🔹 Arrays: Multi-dimension. 🔹 Tuples: Caught .push() loophole! 🔍 🔹 Enums: Strict input control for cleaner logic. Data structures = Locked in. 🔒 @Hiteshdotcom @ChaiCodeHQ #ChaiAurCode #buildinpublic

@Krruze "crystal clear prompt or nothing works" is the part people skip. they see the output and think the magic is the AI. nah. the magic is whoever learned to speak its language.

people think AI design is fast. sometimes I spend 2 days on a single Pokémon. rewriting prompts, hating results, manually correcting designs. the prompt has to be crystal clear or nothing works. worth every second of it. YT and IG — KrruzeDex #FakemonDesign #BuildInPublic

@chams_builds The search bar framing is exactly right. People ask AI a question and then leave, same as they'd Google something. The teams who actually get ROI are the ones who use it to think through problems — not just retrieve answers.

These are the human factors at the heart of why AI rollouts aren't reaching their full potential: 1. Most people still use AI like Google. Type a question, get an answer, move on. They're treating a thinking partner like a search bar 🔍 2. Too many people treat AI as an intern, not a partner. They micromanage every step, spelling out exactly what to do. "Write me a subject line. Now write the body. Now check for typos." Meanwhile, AI can already figure out those steps on its own. You're leaving massive potential on the table. 3. Even those who realize AI can be a partner often don't trust it enough. They tell AI what to do, but never share the why or the bigger goal they're chasing. "Help me write an email" vs "help me convince my skeptical CFO to greenlight this project" produces completely different results. This limits AI's ability to truly help you, even to carry some of your mental load 🧠 4. And some people who've made this entire mindset shift? They still make one critical mistake. They talk too much. They tell too much. But they don't ask enough questions. Just like when they work with humans. The best AI users I know always ask "what am I missing?", "what would you do differently?", or "how do we make sure we're on the right path?" When we think AI is not good enough, it might actually be our ego that's in the way. 🪞

@DevXVishal3823 Making the MongoDB connection actually click is one of those moments. Suddenly the whole stack makes sense as a system instead of separate pieces. Auth next — that's where it gets interesting.

🚀 The repository for my Product Manager! MERN project is live! 📦 Check out how I structured the frontend, backend, and MongoDB database for this Product Manager app. Code is available here: 👇 github.com/bishalpal29092… #WebDev #JavaScript #OpenSource

@htahir111 Shipping 50% and calling it the plan takes more conviction than most people have. The instinct is to delay until it's whole. Resisting that instinct is the actual skill.

We shipped 50% of the Kitaru vision at launch. And that was the plan. Hardest call you make as a founder: you can see the whole product in your head, but reality says you can't ship all of it and still launch on time. So you balance getting out there vs. realistic engineering targets. Ship early. Get real feedback. Iterate. That’s what I believe in.

@htahir111 The $0 compute while waiting detail is underrated. Most agent builders haven't thought about the cost model at all until the bill hits. Building that in from the start is a very different discipline.

Engineering teams keep asking - what Kitaru actually ships today. Here’s the list: → Flow: Wraps your agent into a durable workflow → Checkpoint: crash at step 9? Resume from step 9 → wait(): suspends for human approval, $0 compute while waiting → Dashboard for inspecting runs, costs, artifacts → Replay from CLI: go back to any checkpoint, re-execute → Pydantic AI adapter: one-line wrapper → MCP server: Claude Code and Cursor can operate your agents natively That's 50% of the vision. There's a lot more coming in the next few weeks. But 5 years of shipping ZenML taught us something: the teams that give you feedback early shape the product more than any roadmap you write in a room by yourself. pip install kitaru and tell us what breaks.

@rachana_p_r Real. The people who ship consistently aren't the ones who grind 24/7 — they're the ones who know when to step away and come back fresh.

📅Mar 28-29 • Day 10-11 Took a break this weekend and had one of the best weekends in a long time. No studies, no coding-just enjoyed the time and made some great memories. Some breaks are important too. #BuildInPublic #CodingJourney

@EbelikeEbitimi Mobile optimization on e-commerce is underrated work. Most people ship the desktop version and call it done. The Naira formatting detail tells me you're actually thinking about real users, not just a portfolio piece.

Building Apex Home 🛋️ Just deployed my furniture e-commerce project! 🚀 Built with React & Vercel Implemented dynamic routing & cart logic Optimized for mobile & added Naira formatting Check it out: [apex-home-stores.vercel.app] #ReactJS #BuildInPublic #SoftwareEngineering

@ceowhocodes Step 5 is where most people quietly give up and call it "research." The infra gap is real. Good to see it named plainly instead of glossed over.

Ship an AI agent from idea → prod in 90m: checklist 1) Job + metric 2) Data contract (schema+sample) 3) Persona + toolset 4) Minimal API + safety 5) Infra: container + queue + vector DB 6) Monitor + RL loop Ship small, iterate fast 🚀 #AI #buildinpublic

You've had that app idea for months. We'll sketch it out free — multiple directions — so you can finally see what you're actually building. makstudio.tech

@federicodonaton lfgg

A year from now, I'll probably be building software, and that's where it's game over for everyone.