Kavya (@Kavyabuildss) header
Kavya (@Kavyabuildss) profile picture

🔨 Building SaaS & my MVP studio ⚡ 21, left my first easy job for the thrill of building 💻 Web, iOS & Android Dev 🌍 Shipping, failing, learning, repeating

Joined August 2025
Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)Photo by Kavya (@Kavyabuildss)
7 Harsh Truths About MVPs I Learned Building 30+ in a Year
8
1
26
5,967
wait what
India is playing a whole new game!
15
the scary part of AI payments isn't the AI it's refunds your agent buys the wrong thing. who do you dispute it with? 👀
10
ChatGPT can now spend your money 💳 everyone's asking if that's safe i build payment systems for a living. here's how it actually works 👇
29
Claude Meetings just killed: → "i'll send the doc after the call" → "let me mock that up and get back to you" → the 2 day wait for a proposal → the follow up email nobody reads → "can you share your screen real quick?" → the person whose only job is taking notes → "wait, who was supposed to do that?" → the second meeting to recap the first meeting
Introducing Claude Meetings: add Claude to your team’s video call and it builds docs, whiteboards, and other artifacts with you while you’re still talking. Starting to roll out in beta on Team and Enterprise plans.
1
1
65
we need a trio now
I had to do it 😭😭
27
Fable 5.2 on High ≈ Opus 5.5 on Max at half the cost. medium is the default, switch it up for real work
Introducing Claude Fable 5.2 and Claude Mythos 5.2, designed to tackle the most daunting technical challenges. They’re capable collaborators on ambitious projects in many fields, and show improvements in areas like coding, reasoning, and vision.
1
1
98
made a quick cheat sheet so you don't have to read the whole thread 👇
51
Claude + ChatGPT Images 2.6 🛋️ drag a sofa into the room. it actually sits there furniture websites are about to look very different
76
receipts should feel like paper 🧾
56
Kavya retweeted
show people the view before they pick the seat ✈️
1
41
show people the view before they pick the seat ✈️
1
41
Introducing launch films, painted entirely in code. -> made with Claude Opus 5.5 -> no image models, no stock, no After Effects -> every tree, cloud and frame is editable code -> built in one night Want one for your next launch? 👇
1
101
anthropic yesterday: claude can’t recreate existing video games anymore levelsio today: decompiled battlefield 1942 with claude code to run it in the browser 😭 someone tell him 👀
I may or may have not just decompiled Battlefield 1942's raw C engine with Claude Code to run it in the browser 😬
63
best AI model in the world: 12% on real work tasks 👀 and people keep telling me devs are cooked 😭 the other 88% is literally my job lol
Super interesting Benchmark. GDPx measures how well AI agents handle 206 challenging tasks across 11 economically valuable fields, including chip design, software engineering, finance, and biology. But here's the interesting part: Opus 5.5 wins with 12%, closely followed by Gemini 4 Carbon with 10.9%. But while performing just 1.1% worse, Gemini 4 is about 4x cheaper than Opus 5.5. So cost-performance wise Gemini 4 wins by a lot.
36
designed a bank app where your phone knows about the payment before your card is back in your pocket ⚡
40
own your intelligence 💯 and own your text too. claude now watermarks what it writes and you can’t even check your own stuff 🙃 your model, your keys, your setup. that’s the only way you actually own any of it
Own your intelligence. Anyone that opposes this has something to sell you (perhaps your own intelligence back to you.) Local models need to continue proliferating. The “absolute unit” massive models may be out of reach for now locally, so your best option is to, indeed, keep your options open. Own your harness, data, connections, runtime: your agents. Embrace open source.
1
45
claude now watermarks what it writes. no opt out, even on paid plans universities can check it. media can check it. regulators, even law enforcement can check it you can’t 🙃 and anyone actually cheating just rewrites it with another model and walks away clean so who is this really catching?
154
i still don’t get why people pay for lovable when claude code exists same models. more control. your code lives on your laptop, not on someone’s platform if they raise prices or shut down tomorrow, your app goes with them 🤷‍♀
34
karpathy built his own youtube in 2 hours with claude 🤯 only things he did by hand: api key, server, domain
Weekend hack: HackerTube. I was annoyed with YouTube's recommendations so I had Claude filter down 43M comments over 20 years of HN to 6,454 videos that keep being recommended. *Significantly* better viewing experience, ~2 hours with Opus 5.5: hackertube.org/ - Excellent recs without any logins (state is kept in local browser session) - Thumbnail is a representative frame from middle of the video instead of the super duper A/B test hyper-optimized user uploads - Underlying data from the comment analysis as one zip file, point your Claude at it and ask for your own version - HackerTube Longs ✊ A bit more on the process: I use Claude Code (desktop app for Mac), Opus 5.5 on Medium effort as my default. I only talk to Claude and have it do everything, the only "manual work" I helped with in tandem was in the browser (reminiscent of MenuGen): 1) I had to create some kind of API key from Google Cloud UI to query YouTube for video metadata, 2) I booted up a VPS box for serving and 3) I bought the domain name. Once the box was up I gave Claude the IP and told it update, secure, harden, deploy and optimize the code, and it did the whole thing. Many awesome projects that would have normally taken weeks/months even can now be few-shotted for breakfast. What else are we not thinking of.
1
103
x finally shows if you’re shadowbanned 👀 link’s in the quoted post. you need 10+ posts last month + a 1 yr old account to see it mine: 0 labels. so my reach is just... me 😭
Under the Hood now shows you which posts received a visibility-limiting label, making it easier to see what’s affected: teapawt.org/i/jf/under_the_hood This was a top request — thanks for the feedback. You’ll see the specific posts listed when fewer than 5% of your posts during the report period have that label.
27