Find builders worth following
Useful posts from 6 builders under 10K followers. Search by interest, or browse their work below. Nobody pays to be here.
Your saved searches
Stored only on this device. Choose a search to check the current pool.
Free, no account needed. Run it again whenever you return.
Need a private feed or RSS?
Explore scouting feeds · 14 days free, then €19/month, only if you subscribe. Feeds recheck the collected pool; new matches are not guaranteed.
How these matches were checked
Each dot is one account, lit when Jev has read it: the brighter, the better the match. The numbered ones made the list. Hover or tap any dot to see the account.
Worth reading this week
Posts published in the last seven days, from smaller creators. Read the original and judge for yourself.
- Richelo Killian @RicheloKillian1,033 followers at profile check · 2 Nov 2026
Day 92 for A thing I did not expect about wiring product one into product two: EmailWorx instantly became IsItDisposable's most demanding customer. Every trial user so far poked at the API politely and left. Real production traffic does not poke. It arrives with opinions. Within days of EmailWorx running live, ISID got pushed harder than ninety days of marketing ever managed, and the integration forced a design decision worth sharing. The decision: a trusted mailbox provider layer. When EmailWorx investigates an…
Read original post → - Nick Vasilev @ns_vasilev49 followers at profile check · 30 Oct 2026
Apple’s default pricing ignores purchasing power parity, overpricing apps in countries like Brazil or India. I built a Claude Code skill to automate 175 storefronts: pick a strategy, preview the table, and push to App Store Connect.
Read original post →
How these posts are selected
Each post passes the overall reading threshold and has no core flag above 50% for vagueness, lack of originality, second-hand content or lack of usefulness. We show one per creator. Recent posts appear newest first; archive posts are ordered by their overall reading score, then publication date. These are model judgments, not fact checks. Nobody pays for selection. A small rotating group of accounts is checked nightly; this is not a live view of X.
More from the discovery archive
Older posts from other creators, selected with the same reading rules. Dates below are publication dates; launches, offers and availability may have changed.
- Yann @yanndine8,353 followers at profile check · 29 Jun 2026
Most Claude Code setups cost more than they need to. The difference between a session that costs $50 and one that costs $3 is not the model. It's the routing logic behind it. Most people treat model selection like a one-size decision. Pick the most powerful model. Use it for everything. So they pay full frontier prices for tasks a local model could handle at zero token cost. I spent weeks testing every combination of local, cheap API, and frontier models for one thing only. Building a Claude Code stack that sc…
Read original post → - Tom Benattar @TomBenattar2,178 followers at profile check · 26 Oct 2026
I launched last week. Here's the brutally honest recap of every channel I tried. Google Ads: since October 21, spent €105.90 (€83 in Europe, €23 in the US) for 43 clicks and 2 conversions. Around €53 per conversion. Not great, not terrible for a first week. Meta Ads: between October 16 and 26, spent $416 across two waves (first US only, then US and Europe). 1,545 clicks, 1,208 landing page visits, and 0 conversions. Yes, zero. Traffic that doesn't convert is just an expensive vanity metric. X post: "just bought…
Read original post → - Cristian Deluxe @CristianDeluxe8,442 followers at profile check · 29 Sept 2026
9/ I replayed that whole connect session, then injected the color write. The LED changed immediately. A one-shot write was ignored; the full handshake and config read made it work.
Read original post → - mason @notmason__184 followers at profile check · 31 Aug 2026
your agent's memory now includes the word "no." selvedge v0.3.11: change_type="reject" records paths never taken, prior_attempts reports stated outcomes as exact instead of inferring, and every rejection can carry the condition that would invalidate it. pip install -U selvedge
Read original post →
Collections worth a read
- Where people stop—and where they stayTwo small-app builders describe behaviour their headline numbers miss: a trial that never starts, and a user who keeps returning.Read the collection →
- Two small apps with a specific jobOne makes spending choices visible. The other supports habits that do not fit a daily checklist. Meet the smaller creators behind them.Read the collection →
- Three small-app lessons worth readingThree smaller creators share what they learned about revenue, onboarding and why people use their apps. Read the original posts.Read the collection →
Recently read creators
new or refreshed account readings- @TomBenattarTom Benattar · 2,178 followers71signalFollow on X
- @yanndineYann · 8,353 followers68signalFollow on X
- @ns_vasilevNick Vasilev · 49 followers58signalFollow on X
- @CristianDeluxeCristian Deluxe · 8,442 followers66signalFollow on X
- @RicheloKillianRichelo Killian · 1,033 followers82signalFollow on X
- @notmason__mason · 184 followers62signalFollow on X
Scouting for founders, hires or partners? Keep a search as a private feed that marks new matches in the collected pool. New matches are not guaranteed.
Not in here yet?
Under 10K followers and writing things worth reading? Read your account: with 3 strong posts out of 20, mostly about your own work, it joins Find, and your page shows every search that found you.
Usage preferences
Optional, off by default. Keeps a 30-day marker on this device and sends action counts and whether you returned after 24 hours. No search text or persistent device ID is sent. Untick to remove the marker. Privacy details.