Projects That
Solve Real Problems
(And Get Found)
I build apps, SEO tools, and AI workflows — then make sure search engines and AI engines can actually find them. Mostly for small teams, indie founders, and content shops that need both shipping and findability handled by the same person. Stack-wise: Next.js, Python, n8n, and the unglamorous schema work everyone else skips.
Who's This Booplex Person Anyway?
I've spent 15+ years figuring out why good projects get buried and mediocre ones show up everywhere. That gap between “built something useful” and “people actually find it” is where I live — and it's where code, SEO, and now AI all crash into each other in interesting ways.
When you can build the thing AND understand why Google (and ChatGPT, and Perplexity) aren't showing it to anyone — that's a useful combination. I got here by being curious, breaking a lot of production servers, and reading forums in languages I barely speak at 2am.
Curiosity-driven.
Got a wild idea? Let's turn it into something real. The work I love most starts with 'what if we could...' and ends with a working prototype a few weeks later. Some of my favourite projects came from someone showing me a problem they'd been stuck on for weeks — turned out the answer was a tool nobody had bothered to build yet, and once it existed everyone wondered how they ever managed without it.
Problem-first.
I start with the headache, not the tech stack. The first questions I ask are 'what's actually broken?' and 'who feels it most?' — only then 'what would we build?' Frameworks come and go (and I've ridden plenty of them out), but the muscle of figuring out which problem is worth solving doesn't expire. That's the part I trust myself on after 15 years of getting it wrong before getting it right.
Learning in public.
I document everything — the wins, the spectacular failures, and those 2am 'why won't this work?!' breakdowns. No Instagram-perfect coding journey here. A lot of the posts on this site are essentially me thinking out loud while debugging; people seem to find them useful because they show the messy middle, not just the polished conclusion. If you've ever Googled an error message at 3am and felt seen by a stranger's blog post, that's the energy I'm going for.
“Most days I'm one bug away from giving up and one fix away from feeling like a genius. Ten years in, still both.”
Things I've Built (That People Actually Use)
Here's a collection of projects that solve real problems. No fluff, no “coming soon” — just functional stuff people actually use. Most of these started as personal frustrations I couldn't shake until I'd written enough code to make them go away. A few were built for clients or as open-source experiments; all of them shipped, and the ones still here are the ones still earning their place.
Slopbuster: An AI Humanizer for Prose, Code, and Academic Writing
Developer ToolsMost AI humanizers swap “delve” for “explore” and call it done. Slopbuster studies how AI text actually differs from human — across prose, code, and academic writing — then fixes it in two passes.
Pixelslop: Catch AI Design Slop in a Real Browser
Developer ToolsA design reviewer that opens your real pages in a browser, measures the actual pixels, reads how they land for a human, and catches the AI slop your agent didn't know it shipped.
RunnerKit: Self-Hosted GitHub Actions Installer
Developer ToolsI kept hitting GitHub Actions limits and getting surprise bills, so I built a shell installer that sets up self-hosted runners in minutes instead of hours.
Stitch Kit: Design Skills for AI Agents
AI-Powered35 structured skills that teach Claude Code, Codex, and MCP agents to use Google Stitch instead of generating gray boxes.
pixeltamer: Image Generation That Does What You Asked
Open SourceGenerate, edit, and compose images with gpt-image-2 — bring an API key or use your ChatGPT subscription. Your call how to pay.
bpx-consult: A Council of AI Advisors for Pi
Open SourceLet the cheap model do the typing. Keep a smarter one on the bench for the calls that actually decide how it turns out.
What past clients said
Real words from real people I've actually worked with. Trimmed for length but not for substance — what they said is what they said. The work was custom automation and SEO tooling, which today would be called agent workflows; the problems haven't changed much, only the vocabulary.
“Used his services many times and have never been disappointed with the results. Always delivered on time and on budget.”
Clive Kirkman
SEO automation
“Not only did he respond quickly, work quickly, and deliver the final product exactly as offered — he walked me through the process so I was able to learn from him.”
Joy
Agent workflows
“Couldn't be happier with his services and skill. Recommended him to all my customers and friends.”
Andrew Richardson
Custom web workloads
How I Approach Building
(From 15+ Years of Breaking Things)
15 years of building stuff, watching it break, and figuring out why. No frameworks named after Greek philosophers — just what actually worked.
Full-stack problem solver.
From server setup to user acquisition — I've touched every part of the stack. 15+ years of breaking things in production taught me how all the pieces actually fit together. When the marketing person needs to ship a landing page and the engineer needs to wire an API endpoint, I can hold both conversations and skip the usual handoff debt that costs small teams weeks per quarter.
Growth through reality, not tactics.
15+ years of SEO, social, and marketing taught me what actually moves the needle. Spoiler: it's not the shiny tactics everyone's talking about this week. Most of what I trust came from running my own projects through every bad piece of SEO advice I could find — and then tracking what genuinely moved organic traffic months later, not on day-of when the dashboard refresh feels exciting.
Read my SEO playbookAI as a co-builder, not a buzzword.
I've been working with AI tools since before ChatGPT made it cool. The interesting part isn't that the model can write code — it's how the conversation reshapes how I think about problems. It's less 'AI writes my code' and more 'AI is the first reviewer who challenges every assumption I make' — and that loop, more than the output, is where the value actually lives. The teams I work with get a builder who's spent more hours arguing with language models than most people have spent prompting them.
“Give me one app that solves a real problem over ten that just look pretty in a portfolio. Substance beats style every time — shocking, I know.”
Latest from the Brain Dumps
Notes from the lab — what I'm building, what I broke, what I learned at 2am. Updated whenever a thought refuses to leave me alone.
Schema generation with Claude: every page type, working JSON-LD, zero hallucinations
Schema is the easiest SEO win nobody implements. Here's the Claude prompt + ready-to-copy JSON-LD blocks for every common page type — blog post, product, servic…
I Built a Council of AI Advisors Because the Old One Died Right When I Needed It
My coding agent runs a cheap model to save money. The trouble is the moments a cheap model quietly gets wrong. So I kept a smarter one on the bench — until the …
The complete guide to AI SEO automation in 2026 (the system I actually run)
Twelve weeks of replacing every part of my SEO workflow with AI. This is the full system — keyword research, brief generation, schema, internal linking, monitor…
Got an idea that won't leave you alone?
Tell me about the problem, the project, the weird bug, or just say hi. Every message gets read personally — no autoresponder, no nurture sequence.
Drop me a lineUsually replies within 24–48h, weekends maybe slower.