Vincent Ferreira

I just published Stop The Tester’s Inferiority Complex: QA and Dev Are Equals medium.com/p/stop-the-tes…

Wanna take back the reins of your writing? Force yourself to write. No goals, just pour out whatever is on your mind. Your brain will thank you. I explain that in my article: How I Write Articles Without AI: My Framework for a 100% Human Content medium.com/p/how-i-write-…

I set this up at 4am and spent an hour just talking to it. It's called Open-LLM-VTuber. You get a Live2D animated AI companion that runs completely offline, sees your screen, hears your voice, and never forgets your conversations. The voice interruption system is different from anything I've seen. The AI cannot hear its own TTS output so there is zero feedback loop and zero awkward pauses. It feels like a real conversation. The inner thoughts feature floored me. You see what the AI is thinking as a separate text layer that never gets spoken. You watch the reasoning happen in real time before the words come out. Pet mode puts the avatar on your desktop as a transparent overlay that floats above every window without blocking anything. Drag it anywhere. It follows you. The persona is entirely yours. Import any Live2D model. Write any system prompt. Clone any voice. Swap the entire LLM backend from Ollama to Claude to DeepSeek in a single config line. 100,000+ conversations have already happened inside this repo according to the user reviews. That number is going to keep moving. github.com/Open-LLM-VTube… 6.1K stars. MIT License. 100% Opensource. What would you use this for... co-working, learning, coding, or just chaos?

Bon les mistraliens, on va voir ce que vous avez dans le bide. Moi j'y vais : modèles locaux, abo, API → je teste tout et je vous raconterai. Je vous encourage à faire pareil. Sérieusement. Faut les exciter un peu les copains, leur montrer qu'il y a de la demande. Ils sont capables de rattraper le tir plus vite qu'on ne le pense → les fondations sont là, on l'a déjà vu dans l'analyse de leur BM. C'est pas parce qu'ils sont focus B2B qu'il faut pas y prêter attention. C'est l'inverse. Et un abonné m'a rappelé un truc : le free tier API Mistral (La Plateforme, plan Experiment) c'est 1 milliard de tokens par mois. En clair → de quoi faire tourner un bot, un prototype, ou expérimenter pendant des mois. Sans CB. Sans limite de temps. Oui, par défaut tes données peuvent servir à l'entraînement. Mais tu peux opt-out en un clic dans la console. C'est un choix, pas une condition. Chez OpenAI ? Le free tier n'existe QUE si tu acceptes de partager tes données. Pas de partage = tu paies. Google API Gemini ? Quotas gratuits divisés par 3 à 5 fin 2025. Max 1 000 req/jour. Mistral API ? ~16 000 req/jour en rythme soutenu. Gratuit. 1 milliard de tokens/mois. Opt-out possible. Gratis. Partagez, parlez-en autour de vous. Faut les secouer un peu et leur montrer qu'on est là.

This Post on Reddit was gold A Founder talked to 12 micro saas founders making $5k-$30k/month asked them all one question : how did you actually find your idea ? not one of them said brainstorming here's what they said instead :

Retrouvez tous les skills pour Claude écrits par des développeurs expérimentés ici : Un skill permet à l'agent de savoir exactement comment faire une tâche (utilisation d'api, logique d'utilisation d'un package, audit de code, sécurisation...)

@paolo_scales Great content

LinkedIn completely changed in the past 6 months and you're getting fucked if you haven't noticed Everyone's still posting the same corporate slop wondering why their reach dropped 90% Meanwhile guys who understand the new game are printing distribution Here's what's actually working: The founder-led content shift (corporate pages are dead) Growing a company page from scratch is damn near impossible now Nobody wants to follow some sterile corporate account posting HR speak People follow people. Look at Adam Robinson - 140k followers, turned that into $6M ARR for RB2B in a few months That distribution is impossible to replicate with ads or cold outreach If you're building from scratch, skip the company page entirely Put 100% into your personal account, 3-5x per week minimum Don't even touch the corporate page until you're nailing personal content The content ecosystem play (this is how you scale past the algorithm ceiling) LinkedIn is capacity constrained - you hit diminishing returns after 1 post/day You can't just spam 20 posts like X and keep getting linear results So smart founders aren't running one account, they're running 3-5+ Your founding team, key employees, all posting coordinated content This creates the illusion you're everywhere on the timeline without overwhelming any single audience Here's the progression: - Months 1-3: One founder posting consistently - Months 4-6: Layer in 2-3 co-founders - Month 7+: Incentivize junior employees to start posting Some companies literally pay $500 bonuses for hitting posting cadence Beehiiv does this perfectly - Tyler Denk + entire team active = omnipresent brand pro tip: Use each account to speak to different ICP subsets Dave speaks to marketing leadership, Matt speaks to junior marketers = total market coverage With AI flooding the timeline, you need ways to certify content as human and real Recurring sets are a visual hook that says "this is a real person with a real company" If you have an office, treat it like a film set: - Proper lighting setup - Branded elements (Pylon painted their SF office purple) - Designated person capturing B-roll daily - Behind-the-scenes footage library Posts with relevant media (video/photo) consistently outperform text-only now This isn't optional anymore - it's the difference mbetween looking AI-generated and authentic Content shelf life is up (algorithm showing week-old posts now) Half your feed is content from days or weeks ago LinkedIn tweaked the algorithm to extend shelf life and reduce the content treadmill This means: - Evergreen topics > trendjacking - Your posts get second/third life weeks later -Current events posts die fast when resurfaced Balance your content - trendjacking gets attention but evergreen content has staying power If someone sees your "breaking news" post a week later, it's dead weight Personality is your moat (the only defense against AI slop) ChatGPT and Claude can pump out "pretty okay" LinkedIn posts in 10 seconds The timeline is flooded with generic AI garbage that reads fine but has zero substance What AI cannot replicate: - Your specific metrics and outcomes - Personal anecdotes only you can share - Your actual writing voice and personality - Personal interests that humanize you Actionable plays: 1. Shift from "how to" to "how I" content 2. Share stories only you can tell 3. Case study posts (your own results, with you own system) 4. Use casual tone relative to your industry Another thing you can do is inject personal interests selectively (Gary Vee's blueberry obsession) Your endurance training, reading habits, whatever Humanizes you from "marketing guru" to "real person" Fast-tracks trust and rapport The game COMPLETELY shifted. Founder-led distribution, content ecosystems, visual proof of life, personality moats

@romanbuildsaas GOJI

72% connection acceptance rate. 56% reply rate.33 calls booked in 10 days. Ater 3 months of testing, we cracked the code. We built the ULTIMATE workflow for LinkedIn outreach (100% automated) Want the copy & paste automation template? Reply "GOJI" + follow for access

@burakeregar Can we set RLS to protect from query spamming or something ?

RLS rules are go to approach by AI models... but they shouldn't be. you must do these instead: 1) enable RLS but don't put any rules this will close your database to the outside of the world. after this only service role key can reach your db

☠️ here is why your vibe coded project is not safe when you set db rules, you think you are safe ☣️ watch how i exploit a "safe" supabase db by inserting millions of rows see the thread below to learn how to protect yours (don't just bookmark it, add them to the rules file)

I renamed the app to ScribTrack. It's better for SEO and actually describes what it does. Also working on landing page (still in progress) and the writing prompts system. That's a lot of parallel works but that's how the pieces align 😉 #buildinpublic

@covacut @AnthropicAI @claudeai I agree too!

i'm convinced the most interesting software of the next 5 years won't come from engineers it'll come from designers and artists who finally have the tools to ship the creative artisan arc is here. still early. new project with @anthropicai on @claudeai 🧡

Just wrapped Week 2 on my writing-motivation app. Converted designs → Next.js web app✅ Supabase auth + persistence✅ Smart onboarding that learns your writing style✅ Now the real work begins: finding the angle that makes Mirinkl different from the English-only competition

Thanks to Gemini 3 + Veo 3.1, I was able to animate my mascot while preserving its soft, watercolor style. A few animations like these will be very useful for making onboarding fun and adding more life to certain parts of the app.

I made a mistake with Mirinkl: I asked ChatGPT to list competing apps. The result? An incomplete list. While browsing the App Store, I discovered similar apps... all in English only. Multilingual support + motivational tracking: that's where Mirinkl stands out 🧐

A must-bookmark for vibe-coders. @YCombinator’s guide to making the most of vibe coding:

@benln I have a build in public project just during holidays, it’s a motivation app for writing called mirinkl

Who's building over the holidays? Will add you to a group chat I started on X.