Notes from shipping 11+ apps.
No hype, no listicles. Just what I learned getting real apps through Apple and Google review, building with AI tools that ship, and running a freelance shop people actually trust.
How to make an AI-built website stop looking AI-built
Lovable, Bolt, v0, and the rest leave the same fingerprints — same gradients, same hero, same Stripe-blue card stack. Here's the 9-step pass I run on every project to kill the 'AI smell' without rebuilding.
Read the post →All posts
How to make an AI-built website stop looking AI-built
Lovable, Bolt, v0, and the rest leave the same fingerprints — same gradients, same hero, same Stripe-blue card stack. Here's the 9-step pass I run on every project to kill the 'AI smell' without rebuilding.
Cursor and Claude Code beat Lovable when you actually have to ship
AI builders are fast for prototypes, slow for production. Real editors with Cursor, VS Code, and Claude Code give you the same speed plus the things that actually decide whether you ship — debugging, native APIs, long-running context, and ownership.
Why Loyal Customers Rarely Leave Positive Reviews
Most happy customers stay silent while angry customers write essays. Here's the psychology behind that pattern, the data that proves it, and how to fix your review flow.
App Developer Vetting Checklist: How to Hire Without Getting Scammed
Red flags to catch before hiring. Framework for evaluating developers who ship apps. Checklist + real examples.
The Real Cost of Paid AI APIs: When Self-Hosted Wins
Your AI API bill is climbing fast. Here's the math on when to switch to self-hosted and how much you actually save.
Cursor + Model Surgery: AI-Assisted Coding with Custom Knowledge
Inject domain knowledge into Cursor's AI. Model Surgery workflow for personalized coding assistance.
Free AI Search: SearXNG + Redis Alternative to Paid APIs
Stop paying $5 per 1K searches for OpenAI/Anthropic APIs. Self-host SearXNG + Redis for zero per-search cost. Production-tested architecture with Docker setup.
How to Ship a Web App to Google Play Store: Complete 2024 Guide
Step-by-step guide to take your web app from Lovable, v0, or Bolt to Google Play. Includes screenshots and real timelines.
Local AI That Actually Works: Ollama + Quantization in 2026
Run LLMs locally without the slowdown. Quantization strategies, GPU acceleration, and when local AI actually makes sense.
Model Surgery: How to Transfer Knowledge to LLMs Without Retraining
Transfer knowledge between LLMs without expensive training runs. Alignment scores of 91.7% to 99%+ across scales — here's how it works and why it matters for shipping.
Native vs Web App for Indie Developers: Decision Framework (2024)
When to build native vs web. Cost, speed, complexity compared. Real examples and decision tree.
Building a Data Scraping Stack: Sourcing LLM Training Data
How to ethically scrape web data for LLM training. APIs first, legal framework, cost breakdown.
Shipping AI Tools to the App Store: What They Don't Tell You
Your AI tool got rejected. Here's why—and how to fix it before resubmitting.
How to ship a Lovable app to the App Store (the no-bullshit walkthrough)
A step-by-step playbook for taking a Lovable web app from URL to a real iOS listing — wrappers, signing, screenshots, privacy labels, the whole loop. Written from 11+ shipped apps.
Unrestricted Local AI: Model Surgery for Custom Guardrails
Remove or customize AI safety filters on local models. How it works, when to do it, and the responsibility that comes with it.
Why Cheap Freelancers Can't Ship Apps
You found a $15/hr developer to ship your app. You're about to learn why that costs way more than $399.
Why Cursor Can't Ship to App Store
Cursor writes code. It doesn't sign certificates, handle provisioning, or submit to App Store. Here's the gap.
Why Lovable Apps Get Rejected From the App Store
You built with Lovable. Now Apple rejected it. Here's why—and exactly how to fix it.
Why Offshore Developers Struggle With App Shipping
Timezone delay is the killer. A 13-hour gap turns a 30-minute problem into a 3-day ordeal.
Why Slapping AI on Your Site Doesn't Work
You added a chatbot. No one used it. You added recommendations. Zero conversions. Here's why bolting AI on fails.
Why v0 Apps Don't Scale
v0 shipped your MVP in 2 weeks. Now you have 1,000 users and it's slow. Here's why and what to do.
Why Wrapping a Web App Doesn't Work (Anymore)
Wrapping your web app in Capacitor or Expo used to work. It doesn't. Here's why, and what actually works.
Why You Should Never Hire for App Shipping (Without Vetting)
Not all developers are equal. Here are the 5 questions that separate who's shipped from who'll waste your time.
Why Your AI Chatbot Seems Dumb
You added an AI chatbot and it hallucinates. Here's why and what actually makes it smart.
Why Your App Gets Rejected for Guideline 4.2
Guideline 4.2 is Apple's most common rejection. Here's what it actually means and how to avoid it.
Why Your Firebase Bill Exploded
Month 1: $5. Month 2: $300. Month 3: unlimited. Firebase pricing is a trap.
Lovable vs. Bolt vs. v0: which AI builder is actually shipping-ready?
I've shipped real apps from all three. Here's what each one's actually good at, where they fall over, and which I'd pick for a project that has to ship to the App Store next month.
Why you shouldn't hire a random freelancer for app shipping (and why you should hire someone like me)
Most app-shipping projects fail at handoff, not at code. Here's how I vet myself the way I'd want a client to vet me — and what to look for in anyone else you're considering.
Apple Guideline 4.2 rejection: the appeal letter that actually works
If your app got hit with Guideline 4.2 (Minimum Functionality), the binary is probably fine — your appeal isn't. Here's the letter I use, why each line is there, and what to change in the build itself.
Add AI chat to your website in one afternoon (Lovable + Supabase walkthrough)
A working AI chat widget on a real site, set up in a few hours. Real code, real costs, real tradeoffs — and why most off-the-shelf chatbots are useless without your data.
Pillars
Real walkthroughs for getting iOS and Android builds approved.
Building shippable web apps with Lovable, Bolt, v0, Cursor.
Adding AI chat, agents, and automations to existing products.
Vetting freelancers, fixed-price vs hourly, and what reviews really mean.
Real projects — what shipped, what almost didn't, what saved them.
Sharper questions, smaller answers — the long-tail of shipping work.