Tulsi Prasad ⚡️

7.9K posts

Tulsi Prasad ⚡️ banner
Tulsi Prasad ⚡️

Tulsi Prasad ⚡️

@thebuildguy

Building $1M ARR in SaaS 🚀 | AI agents that work while you sleep 🤖 | Shipping daily, failing forward | Documenting the my ups and down so you don't have to 💪

Remote Katılım Haziran 2015
430 Takip Edilen798 Takipçiler
Sabitlenmiş Tweet
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
I got tired of manually copying env vars between 8 different AI coding tools and researching for hours for the exact config format then end up messing everything else. 🚨 Hence, I built Confkit — Never spend a minute on MCP config again! ✅ Auto-sync all MCP clients (codex, cc, gemini etc) ✅ One click MCP addition to all environments ✅ Profile switching for handling multiple env with unique MCPs/envs No more copy-pasting API keys at 3am. #buildinpublic #indiehacker #DevTools
English
8
2
38
2.8K
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
@VadimStrizheus Yeah it appeared on the /model list earlier not sure why it doesn’t show up anymore. Worked really well
English
0
0
0
58
Vadim
Vadim@VadimStrizheus·
HOW AM I JUST FINDING OUT ABOUT THIS?! You can use /model opusplan Claude Code uses Opus for planning and Sonnet for execution. This is effiencymaxxing👇
English
67
52
1.4K
96.9K
Tulsi Prasad ⚡️ retweetledi
Palantir
Palantir@PalantirTech·
We’ve launched a 1-year internship program. For college students looking to make your mark on the real world right now, we recommend you apply. Good luck.
English
640
609
16.5K
3.9M
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
Social media scheduling platform is the next version of todo list app :)
Tulsi Prasad ⚡️ tweet media
English
1
0
6
60
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
Really love the trial pricing here - on @postel_app 🔥 Especially as it's hard to stay off from freeloaders but it's better to giving a (paid) trial to make sense of the app before commiting full-time!
Tulsi Prasad ⚡️ tweet media
English
2
0
2
53
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
Hey @AnthropicAI the recent updates on claude code web are really awesome! Absolutely loving it! But why does it not allow to upload any other formats of files like CSV, XLSX etc? This is restricting to give a pure local like experience!
Tulsi Prasad ⚡️ tweet media
English
1
0
5
90
punarv
punarv@ycocerious·
What do you guys do while your claude code is running?
English
501
2
293
58.9K
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
Hey @aquavoice there's a weird bug on the app or maybe I'm using it wrong. Sometimes I just press my hotkey to start recording speak for like few minutes then when I again press the hotkey to stop recording and paste transcript, it just disappears. No history, no saved voice or anything. So my last 2 mins of prompting gets wasted and I have been facing it since months (<5% of times), but I don't know if it's a bug or any config I need to use.
Tulsi Prasad ⚡️ tweet media
English
0
0
0
59
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
@alexhillman How does it work consistently? Do you have to remind it always or is it stored in its TOOLS or you have made a skill for it?
English
0
0
1
1.7K
📙 Alex Hillman
📙 Alex Hillman@alexhillman·
People really are sleeping on the amount of UI that you can give Claude Code by giving it access to Discord.js and letting it make buttons, modals, and more.
English
4
2
80
11.7K
Tibo
Tibo@tibo_maker·
ok the secret is PG is my business coach for the last 10 years every time I'm stuck on a product issue, a team decision, or whatever I go back to him (through his essays 😅 ) & find the answers these are the 10 essays (out of 231) I re-read at least once a year: 1. do things that don't scale the best founders don't wait for systems. they do things by hand, one by one, doing whatever it takes I still do this - every new product, I'm the first customer support rep. you learn more in a week of manual work than a month of dashboards. 2. how to start a startup three things kill most startups: the wrong people, building something nobody wants, and running out of money. fix all 3 & you'll probably survive sounds obvious - but most people don't do it 3. how to get startup ideas the best ideas come from living at the edge of a problem, not from trying to brainstorm your way to a billion-dollar market every product I've built started with something that annoyed me personally - not a market research doc 4. maker's schedule, manager's schedule makers need half-day blocks to do real work. a single interruption doesn't cost 30 mins, it costs the whole morning I took it further. no meetings at all - everything async on Slack 5. startup = growth a startup is defined by only one thing: being designed to grow fast. if you're not growing fast, you're a business, not a startup most people confuse the two. both are fine, but they require completely different decisions 6. how to do great work great work sits at the intersection of what you're naturally good at, what you're genuinely obsessed with & where there's real room to do something ambitious it's not about startups at all. it's about not wasting your life 7. the 18 mistakes that kill startups every way a startup dies traces back to the same root: not building something people actually want I use this as a checklist before launching anything. I've personally made at least half the mistakes on the list. some of them even twice. 8. founder mode the advice "hire good people & give them room" works for professional managers. it's often the wrong playbook for founders, who need to stay close to the details to keep the company alive standard management advice nearly broke two of my products. this essay made me feel less crazy for ignoring it. 9. life is short life being short isn't just a figure of speech. it's a filter. if something isn't worth the finite time you have, you should cut it this one isn't about startups, it's about everything else. I think about it when I'm about to say yes to something that doesn't actually matter 10. how to make wealth wealth isn't money. money is just how wealth moves around. a startup is a way to create real value for people, not extract it this one shifted my whole frame. building isn't about capturing a slice of something. it's about making something that didn't exist before -- it's the most valuable startup education on the internet
Tibo tweet media
English
52
13
297
25.6K
Anubhav
Anubhav@Anubhavhing·
Crawling an entire website used to take: A Python script. Playwright or Selenium. Proxy rotation. Rate limiting logic. Error handling. 3 hours of debugging why page 47 returned a 403. Now it's one API call. Every web scraping startup that raised millions to solve this problem just became a single endpoint. Every freelancer charging $500 to "extract website data" just lost their entire business model to a /crawl command. HTML. Markdown. JSON. Pick your format. No scripts. No browser. No headache. The entire web scraping industry just got reduced to one line of code. Someone is going to use this to clone every competitor's website by Friday. 💀
Cloudflare Developers@CloudflareDev

Introducing the new /crawl endpoint - one API call and an entire site crawled. No scripts. No browser management. Just the content in HTML, Markdown, or JSON.

English
229
512
9K
2M
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
@mattwelter At this point I don’t even know if I can call it work, if it feels so much like play 😭
English
0
0
1
31
Matt Welter
Matt Welter@mattwelter·
am i a workaholic? i said i'll take a break saturday worked half the day i said i'll take a break sunday worked half the day again i said i'll take a break after 6pm did that for 2 days it's 10:30pm and i deployed 5 new features in the last hour i haven't taken a day off in months am i a workaholic?
Matt Welter tweet media
English
24
0
55
3.8K
Anish Moonka
Anish Moonka@AnishA_Moonka·
India's own government think tank put a number on this. NITI Aayog's October 2025 report says that if nothing changes, the tech services workforce drops from 8 million to 6 million by 2031. Customer service goes from 2.5 million to 1.8 million. That's 2.7 million jobs gone from just two sectors in six years. But the same report shows the opposite outcome is just as plausible. If India moves fast on AI skilling, the tech workforce could grow to 10 million, with 4 million new AI-native roles by 2030. The gap between losing 2 million jobs and gaining 4 million comes down to what happens in the next 3 to 4 years. WEF projects 92 million jobs displaced worldwide by 2030, but 170 million created, a net gain of 78 million. In India specifically, EY forecasts that 38 million jobs will be transformed by GenAI by 2030. ServiceNow estimates agentic AI will redefine 10.35 million Indian roles in that same window, with manufacturing taking the biggest hit at 8 million. Over 60% of formal-sector jobs in India are expected to be susceptible to automation by 2030, according to NITI Aayog. The most exposed roles: L1 customer support, QA testers, data-entry operators, and routine coding. Entry-level BPO jobs that once absorbed millions of graduates are already drying up. Annual hiring growth in business process management has fallen from tens of thousands to under 20,000. The pipeline math makes this personal. India produces 1.5 million engineering graduates a year. In 2022, IT companies hired 600,000 of them. By 2024, that collapsed to 70,000. AI talent demand in India is growing at 25% annually, but the supply of AI-skilled workers grows at only 15%. NITI Aayog flagged that India produces fewer than 500 AI PhDs per year. Its share of global AI patents has dropped to under 5%. Reliance committed $115 billion over seven years for AI infrastructure starting in 2026. India's AI Mission has deployed 38,000 GPUs, triple the original target. NASSCOM projects 1 million new AI-related roles. Those roles need people who can fill them, and right now, only 45% of Indian graduates are employable in the digital economy, according to NASSCOM's own data. "Millions of jobs disappearing" is half the story. The other half is millions of new jobs appearing that most of the current workforce can't fill yet. What India is staring at is a replacement cycle in which the entrance to the new economy is blocked while the exit from the old one is accelerating.
English
15
103
1.2K
175.9K
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
@leonabboud Glad to see you doing this, there's never a one-size-fits-all, and then they end up with more questions than answers. 🤦‍♂️
English
0
0
0
12
Leon Abboud
Leon Abboud@leonabboud·
The only way to deal with tire kicker leads is by holding your frame. Had this lead book a call from a YouTube video. Right after booking, he sent an email asking if I could send over a proposal via email. Said no. Most people would cave and send out random decks to prospects who aren't even serious. If this ever happens to you, hold your ground. This is your sales funnel, you follow your sales process, not that of the leads. You will lose a few leads that way, but remember, there are millions of fish in the sea. Losing an unserious lead makes space for a serious one.
Leon Abboud tweet media
English
17
0
49
1.9K
kitze 🛠️ tinkerer.club
breaking news: person who's job is to write "continue pls" and "fix pls" 49 times per day thinks that they're gonna get overpaid to keep doing this task forever more news at 11
English
21
8
563
35.1K
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
Okay, it finally worked. I didn't know what worked, login via GCP or Google, it randomly took me forward 😂 But the design it (Gemini 3.1 Pro) did was pretty decent, I'd say better than Claude Opus 4.6. In claude's version even tailwind setup was broken. (yes with the frontend design skill) POC and vision shown to the client, they liked it. Let's see how it goes. But man the designs by goat @MengTo with gemini, it's just too hard to do by just one shot prompting!
Tulsi Prasad ⚡️ tweet mediaTulsi Prasad ⚡️ tweet media
Tulsi Prasad ⚡️@thebuildguy

YES I'm trying Antigravity today (built by the giant 400BN ARR Google), but I'm stuck in their onboarding since last 15 mins can't go forward. Yes the market is all yours, any startup you can kill with 1 liner in your I/O but killing the user experience is not gonna work.

English
1
0
3
204
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
I've been coding non-stop for last 24 hours but best part is, it's all happening through Claude Code Web. Why CC Web? 1. Web means you can literally spawn as many agents as you want for different things (bugs, features etc) 2. I tried Codex cloud earlier, but in actual bigger applications, with separate frontend, backend, infra repo's it's hard in web to give context for different repos. But in CC web you can literally add more repo's into context. How do you review your code? 1. Setup live preview deployments, so anywhere you're from, you can push updates and in Github PR preview check it out. 2. More PR reviewer tools like CodeRabbit, Gemini and the best reviewer Codex, do help a ton if you're shipping insanely fast but want to uphold the quality. The future of coding is changing, but love the direction it's going into giving more of a systems thinker and a PM role instead of manual hardcoding labour.
Tulsi Prasad ⚡️ tweet media
English
0
0
5
73
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
@robinebers How is Gemini 3.1 Pro compared to that for design work? I just installed Antigravity let's see how it goes
English
1
0
1
166
Robin Ebers | AI Coach for Founders
using Opus 4.6 a lot today to design a new landing page and holy shit is it frustrating to work with such a dumb model constant babysitting, so much wasted time HATE ME I DON'T CARE I will preach it until you wake up if GPT 5.4 was better at designing/writing, I wouldn't kill my sub TODAY
Robin Ebers | AI Coach for Founders tweet media
English
40
2
42
5.5K
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
YES I'm trying Antigravity today (built by the giant 400BN ARR Google), but I'm stuck in their onboarding since last 15 mins can't go forward. Yes the market is all yours, any startup you can kill with 1 liner in your I/O but killing the user experience is not gonna work.
Tulsi Prasad ⚡️ tweet media
English
10
0
14
522
Tulsi Prasad ⚡️
Tulsi Prasad ⚡️@thebuildguy·
@boringmarketer The best part is everyone having access to the same tools, makes us value the most important things, the most!
English
0
0
0
8
The Boring Marketer
The Boring Marketer@boringmarketer·
with ai execution is abundant. judgement and taste are not. the bottleneck is not "can it be produced" the bottleneck is: • what matters • what should be done next • why do I do it • what is worth automating • what is most likely to work • what learnings are feeding the system the best ai operators are becoming systems thinkers
English
29
2
100
6.7K