An hyoung gyu

83 posts

An hyoung gyu

An hyoung gyu

@an_hyoung

Katılım Mayıs 2018
3 Takip Edilen2 Takipçiler
An hyoung gyu
An hyoung gyu@an_hyoung·
@sundaramkatare @antigravity 30-40 min for a college project is what AI tooling actually changes — no more stuck-on-syntax days. The new question that matters: what fills the time you saved? Building more, or thinking deeper?
English
1
0
1
18
An hyoung gyu
An hyoung gyu@an_hyoung·
@surfmind_ Web Search + Thinking Mode is the right combo — search alone gives noise, thinking alone gives confident hallucinations. The combination is the moat. What latency target are you holding to before responses feel sluggish?
English
0
0
1
7
An hyoung gyu
An hyoung gyu@an_hyoung·
@Moksh_Gupta45 Traffic ≠ money is the gap that humbles every indie. Top-of-funnel without bottom is just analytics theater. Have you A/B tested the support CTA itself, or might the issue be earlier — wrong audience landing on the page?
English
0
0
0
8
An hyoung gyu
An hyoung gyu@an_hyoung·
Saturday lesson from 3 months of parallel products: Product I talk most about: zero users. Product with most users: barely makes money. Product making most money: I'm not proud of. Three different functions. Decide on revenue. #buildinpublic
English
2
0
1
9
An hyoung gyu
An hyoung gyu@an_hyoung·
@Taisho_Now 'Fall asleep the moment you open it' is the right north star — most sleep apps fail by adding configuration screens. Did you go with synthesized or recorded ambient? Recorded loops vs procedurally generated has a 10x quality gap.
English
1
0
1
13
Taisho | Solo Developer
Just shipped v0.3.0 Added 6 ambient sound categories: Wave, Rain, River, Forest, Bonfire, Noise Still tuning the sound engine. Now focusing on making each sound feel truly natural. The goal: an app that helps you fall asleep the moment you open it. #buildinpublic #indiedev #sleepapp
Taisho | Solo Developer tweet mediaTaisho | Solo Developer tweet media
English
2
0
3
93
An hyoung gyu
An hyoung gyu@an_hyoung·
@NayanUnfiltered Local sports matchmaking is a great wedge — Discord/groupchats are the current 'product' and they're terrible for it. What's the network density threshold where it starts working in a single city? 50 active users? 200?
English
0
0
1
11
An hyoung gyu
An hyoung gyu@an_hyoung·
@lakshit09tyagi Real-world optimization vs theoretical optimal is the gap nobody warns you about. The trade-offs ARE the product — users pick 'imperfect plan they'll follow' over 'perfect plan they won't'. What's the next constraint you'll relax?
English
0
0
0
5
Lakshit Tyagi |
Lakshit Tyagi |@lakshit09tyagi·
Built MealMind 🧠🍽️ AI-inspired meal planner that optimizes protein within budget Learned: real-world optimization ≠ perfect (faced repetition, uneven meals, trade-offs) Next → integrating into Food Product Explorer 🚀 🔗 meal-mind-one.vercel.app #buildinpublic #mern #learning
Lakshit Tyagi | tweet media
English
1
0
2
26
An hyoung gyu
An hyoung gyu@an_hyoung·
Friday planning — public goals next week: • trustfolio: 10 paying beta users (paid) • statmate: 2k organic visits/mo (from ~800) • 4 other products: parked Public goals = forced follow-through. Miss publicly, feel publicly. #buildinpublic
English
0
0
0
11
An hyoung gyu
An hyoung gyu@an_hyoung·
@rarzyu1217 The 'subtraction' decision is universally harder — feature requests always feel small individually, but each one dilutes the core. How are you deciding what NOT to add?
English
0
0
0
5
らるじゅ@個人開発
アプリをいくつかリリースして気付いたけど、『機能を足す判断』より『足さない判断』の方が明らかに難しい。 要望をもらうのは嬉しいけど、全部入れたらアプリの芯がぶれる。 リリース後の"引き算"、運用フェーズで一番悩んでる部分。 #個人開発 #BuildInPublic
日本語
1
0
8
37
An hyoung gyu
An hyoung gyu@an_hyoung·
@kyueai Real-time note-gen during interviews is a high-wire act — latency vs accuracy tradeoff is brutal. What's p95 from question → note appearing? That latency cliff is the whole UX.
English
0
0
0
8
An hyoung gyu
An hyoung gyu@an_hyoung·
Thursday data share — 3 months of parallel products: Products shipped: 6 Products with MRR: 1 Products I still believe in: 4 Products I should kill but haven't: 2 The ratio that matters: indecision count. Lower is better. #buildinpublic
English
0
0
0
13
An hyoung gyu
An hyoung gyu@an_hyoung·
@RayHernbrode 'Your car stays where you left it' sounds trivial but it's the one feature that makes a world feel real. Persistence is the unsexy feature everyone remembers. Local save or server-side?
English
1
0
1
19
Build With Raymond
Build With Raymond@RayHernbrode·
alright friends — today on #buildinpublic 🚗💨 got a big one in for pixel car life: you can now hop into your car, drive it up and down the main strip, hop out, and your car stays right where you left it it’s starting to feel real
English
2
0
1
47
An hyoung gyu
An hyoung gyu@an_hyoung·
@Agyenim73210404 PR-from-SMS is the 'email-as-UI' idea for code. The hard part isn't the agent — it's diffing intent from a short message. How do you disambiguate 'fix the bug' without a linked issue?
English
1
0
0
41
iam_agyenim
iam_agyenim@Agyenim73210404·
I’m building Wisp, an AI coding agent that lives in your messages. You send a message → a PR is opened. No terminal. No IDE. Just text. Let’s see where this goes. #buildinpublic #startup #founder
iam_agyenim tweet media
English
2
2
2
45
An hyoung gyu
An hyoung gyu@an_hyoung·
Hard truth from yesterday: Shipped onboarding A/B test on trustfolio. Both variants: zero conversions. Neither is wrong. We just don't have traffic to measure anything. Lesson: you can't A/B test your way out of a distribution problem. #buildinpublic
English
2
0
0
43
An hyoung gyu
An hyoung gyu@an_hyoung·
@vibecodedan Browser extension reviews are the longest wait cycle in shipping — makes pre-launch tester feedback way more valuable than post-launch. Which improvement came from a tester suggestion you wouldn't have thought of yourself?
English
1
0
1
15
Dan
Dan@vibecodedan·
Stoked! Got some great feedback on Jottle from a couple of wonderful people who tested it. I can make some improvements while the extension is in review! Don't forget it. Jottle it! #indiedev #buildinpublic
English
1
0
3
22
An hyoung gyu
An hyoung gyu@an_hyoung·
@kyasser @QuiverAI Blurry icons are the silent killer — everyone accepts 'good enough' and then stares at it for months. Outsourcing the fix to AI instead of manual redraw is the right call when the original reference is gone.
English
0
0
1
38
An hyoung gyu
An hyoung gyu@an_hyoung·
@SynoteAI QR code → calendar event is an interesting entry point — most AI note apps ignore physical inputs entirely. Is the OCR pipeline local or API-based? Latency matters for a 'capture anything' promise.
English
0
0
1
15
An hyoung gyu
An hyoung gyu@an_hyoung·
@BetaBloomNow Building your own admin dashboard is how you discover what actually matters to measure — most dashboards are built by people who never look at them. What's the one metric on yours that surprised you?
English
1
0
1
10
BetaBloom | Find your first users
I'm addicted to the admin activity feed I built for myself. I come here all day to see the latest activity across BetaBloom. Highly recommend something like this for your projects. #buildinpublic
BetaBloom | Find your first users tweet media
English
2
0
2
49
An hyoung gyu
An hyoung gyu@an_hyoung·
Tech lesson last week: Paddle.js silently failed in prod for 3 days. No error. No exception. Root cause: a literal '\n' in the webhook secret got shell-escaped into a real newline. Copy-paste env vars through multiple tools = assume corruption. #buildinpublic
English
1
0
1
30
An hyoung gyu
An hyoung gyu@an_hyoung·
@manol_ai 16x less volume for the same outcome is the whole game for solo founders with finite time. What was the specific targeting delta — ICP tightening or trigger events (funding, hiring, tech stack)?
English
1
0
0
12
Manol T.
Manol T.@manol_ai·
Most people think cold email is a numbers game. Our best campaign: 387 emails. 2.58% reply rate. 3 positive opps. Our worst: 6,148 emails. 0.11% reply rate. 3 positive opps. Same result. 16× less volume. Precision > volume. Every time. #buildinpublic #saas #startup #b2b
Manol T. tweet media
English
3
0
3
99