I keep a notebook of app ideas, and it isn't a junk drawer. It's an index file, tier-ranked S+ / S / A / B, where the best ideas already carry a name, a one-line pitch, brand colors, an MVP scope, competitor research, and a marketing hook before I've written a line of code. The discipline isn't having ideas. It's knowing which few are worth a year of your one life — and being able to throw the rest away without ego.
I'll be honest about the counterweight up front, because it's the most me thing here. My gift is being fast at finding the right thing to work on. My flaw is the opposite of that.
“The actual bottleneck is finishing, not picking. I'm fast at finding the right thing and slow at completing it.”— from the notes
So read this as what it is: a menu, not a résumé. A few of these shipped (those live over on The Maker). Most are notes — sharp, scoped, waiting. The whole reason this page is public is the same reason the rest of the site is: I'd love for one of these to become a we. If any of them is your kind of problem, grab it — I'm far better at this with someone who's better at finishing than I am.
A note on names: most of these went through a naming round before anything else. I find the pitch and the App Store page at the same time as the mechanic. So the bold name is usually the real working name.
The top shelf (my S+ / S picks)
These are the ones I'd actually pick up next. They've earned the top tier because I keep coming back to them — and because each one is the same engine pointed at a problem I personally have.
- Friend Tracker (top pick) — the app that reminds you who you haven't talked to and nudges you to reach out, paced by how close you are. "Don't let your friendships drift apart." You rate each friend on a 1–10 closeness scale; your best friends should hear from you weekly. There's a little avatar whose health tracks your social life, a shared note-space per person ("ask how their YouTube channel's going"), birthdays and weddings, and a travel map so you can catch up with people in a new city. This one isn't a clever toy — it's the thing I most wish existed, built straight out of my deepest value.
- RANKED Language Learning (top pick) — five-minute live 1v1 ranked language battles, with an ELO that auto-calibrates question difficulty and negative marking so guessing actually costs you. Think competitive comprehension-speed. It's meant to fold into Flua as a ranked mode and ride a streamer funnel ("long-time viewer made a ranked languages app").
- StoryWeb — Cursor for writers (top pick) — an infinitely-zoomable card board for novelists: plot beats → chapters → scenes → beats → manuscript, keyed 1–5, with parallel linked boards for character arcs, a reader-knowledge / reveal map, and worldbuilding. A git-diff-style relationship graph you scroll through story-time, an audit mode that flags "dangling" cards, and an AI "story refactor" tier on top. Explicitly inspired by clean software architecture — the most fully-spec'd idea in the whole notebook, and the exact place my engineer-self and my writer-self turn out to be one person.
- Phone App for a Cursor-style IDE — git-connect to the project on your laptop, talk to an agent that edits and runs the code on your running machine, and get a GitHub PR link to review — all from your phone. So I can build on a walk.
- Workout Tracker & Planner — scores how complete a workout is per muscle using real recovery math (bench = 30 chest points; hitting the same muscle on different days gets a rest bonus; back-to-back stacks badly). Shareable plans. The science version of the body-as-a-build.
- Drumroll Timer — a board-game turn timer where a drumroll builds as the clock nears zero, single-tap to reset, board-game-themed throughout. Tiny, tactile, done-in-a-weekend — the kind of single-purpose thing that just has to feel perfect.
Best-of-both-worlds media remixers
The recurring move here is taking two things people think you have to choose between and inventing the thing in the middle — and, yes, loudly claiming I invented it.
- SubStepper (standout) — turn anime into a visual novel. Instead of seeking ±10 seconds, you seek to the next or previous subtitle line, so you step through a show at your own pace like a manga — skimming slow scenes, letting the big ones simmer. A variable-speed mode speeds up only the non-dialogue parts so the audio stays perfectly natural. The pitch wall writes itself, and I went a little feral on it:
“I SOLVED the anime vs manga debate by inventing this tool which gives you the best of both worlds.”— SubStepper notes
- “seeking +-10seconds in STUPID” — the entire product thesis in four words
- “You can watch ONE PIECE in 100hrs~ — no manga, no one pace, no 2x → best method”
- Honest blocker, written right next to the hype: anti-piracy baked-in subtitles make it hard to build, so first I'd test it in a playground and pitch it to someone
- Subtitle Surfers — the throwaway sibling of SubStepper, filed under "brainrotmaxxing while watching Netflix." Same seek-by-subtitle idea, lower brow, fully self-aware.
- AI Infinite-Idea Generator (standout — this is the engine, as an app) — an endless scrollable wall of AI-generated ideas (book covers and titles, YouTube concepts) where scrolling toward a niche surfaces more of it, backed by DB-cached exploration so you're not re-generating the same wall. The design note I care most about: "make scrolling feel tactile and meaningful so they don't just single-swipe through 100 ideas." It's my idea engine, turned into a product.
- Audiobooks You Can Talk To — an AI narrates in a consistent voice, stops when you ask a question, answers in-character, and resumes — so it feels like one entity is telling you the whole story. My own note flags the obvious: "very high perceived authority (can be used for wrong)." Filed honestly, including the catch.
Tools for tables — D&D and board games
This whole cluster is the games half of me cashing out as software. I DM, I've sleeved 40+ board games, and almost all of these come from a real annoyance at a real table.
- D&D DM Live-Assistant (standout) — it listens to the session live and pre-pulls what you're about to need, the instant before you need it. A player talks to a random NPC and a one-liner appears; you say the line and it's replaced by deeper notes, rumors the NPC might drop, the next hook. The design goal in one sentence: as soon as the DM would need to check the PDF or Google something, it's already on screen.
- YouTube-Based D&D Sounds — "tmux for YouTube": preset ambience stages, one-tap fade between them, auto-loop, and a save-by-vibe-tag library. A copyright-free version on Epidemic Sound. A better, cleaner Syrinscape.
- Board Game Assistant & Avalon / Werewolf Roll-Assigner — play party and board games with no physical components; everyone's phone joins one live-synced session for secret role assignment and the rest.
- Arkham Horror Mobile — play the Arkham Horror card game on mobile, wired to the existing card database. My note: "straightforward to execute with what I have now."
- Better Splendor — a Splendor app where the card movement feels incredible: cards that lift as you drag rather than the instant you touch them. The whole value prop is that every single move feels good, and grabbing high-value cards feels great.
Minimal, feel-good utilities
Small, single-purpose, design-obsessed apps that exist mostly to feel perfect. The whole pleasure here is making one tiny thing flawless.
- Kalden's Minimalist Metronome — because "no metronome app looks good, minimal, and isn't buggy." Dead simple, haptic, fully free, and a quiet funnel to my other apps.
- My Minimal Obsidian Setup / Obsidian Mobile Extension — a wrapper that pre-installs the good plugins and overhauls the mobile UI (one bottom bar, swipe between notes, a typewriter theme). The hook: a million downloads and nobody's done a proper mobile UI overhaul.
- Hour Log Work Tracker — a minimal time tracker by activity — writing hours per week, reading, socializing — laid out like a calendar so you can see where your life actually goes.
- Vibe Pitch Trainer — vibration-based singing feedback: it plays a random note, you sing it back, rolling pitch detection scores you. The utility version of the singing practice.
- Novel Budgeting App — budgeting that treats one-off "things I want to buy" as a ranked wishlist alongside the monthly budget, and gives self-investment an aggressive allowance that steals from other categories. The stuff you want most, you actually get; the rest quietly demotes itself until you stop wanting it.
The wider catalog
The lower tiers aren't lesser ideas so much as further-from-me ones — but the range is the point, so here's an honest slice. The moral spread is real, too: some of these are civic, some are cheerfully cynical, and I've left both in.
- Game Picker — "What do I play?" — a decision tree that buckets your Steam library by mood and craving rather than genre, by remapping Steam tags into emotion buckets. Sign in, filter, get told what to play. One of the most fully-engineered specs in the notebook — it ends with me asking for the whole Flutter codebase, TDD and clean architecture.
- PillKart — Pill Scheduling — reminders, multi-pill inventory, and a prescription/photo scanner via a vision model — packaged name-and-logo-first, with an "Alpha 0.2" that's an actual physical pill dispenser. The most complete example of my brand → MVP → competitor-research ritual.
- Fastag Dashcam Auto-Reporting (the civic one) — hit a button to save the last 60 seconds, an AI reads the license plate, and it files a report; crowd validation plus a 5-star driver rating, with device-signed timestamps so AI fakes can't forge evidence. Road safety as a product.
- Video Tinder — a shorts-style feed of dating videos, so personality and voice carry the match instead of photos. And the growth-mindset footnote I left myself, which I love: "Dammit someone already did exactly this — Snack. Although I can't seem to download it, maybe it died — if so we can try again." Ships anyway.
- City Explorer by Neighborhoods — explore a new (or your own) city neighborhood by neighborhood, auto-checking-in by location, with completion percentages, percentiles, per-category coverage (cafés, attractions), user-curated "vibe" photos, and safety markers.
- Modular Stat Tracker — a renamable, template-driven tracker that's "D&D Beyond for any RPG" and a workout log and anything else, graphing any combination of stats over time or over their own changes.
- D&D Beyond for Other Systems — the same character-sheet tooling for D20, Nimble, and Pathfinder, with playtime conversion so you can bring your 5e character into a Nimble game.
- Storyboard → Manga AI Generator — turn a rough sketch storyboard into a finished manga or webtoon page, so a beginner can actually make one. An easy AI wrapper now that the models are good enough.
- Collaborative Drawing Lobbies — Procreate-style rooms with video-game matchmaking, so you're not alone while you learn to draw — find an active lobby and join it, maybe with a voice call built in.
- YouTube Takeaway Summarizer — a browser extension that gives you the one-line answer to the question a video's thumbnail is asking, so you don't have to watch it.
- Tipping UI Prank App & "Just Fucking Tetris, No Ads" — yes, those are in there too. The notebook contains jokes, and I keep them, because the honest version of an idea engine includes the dumb ones.
How the engine actually runs
The economics are why I rank this hard instead of building everything. My blunt working model:
“Apps will make 0 money or infinite money. There is no in between.”— App Earnings Estimate
If an app hits a real market and is genuinely the best at one thing, the stores and word-of-mouth do the rest past a few hundred real users. If it doesn't, no amount of pushing saves it. That's exactly why the discipline matters — the ranking is me trying to spend my scarce finishing on the one or two that could actually go infinite, instead of spreading it across fifty that go to zero.
And the deeper reason all of this is sitting out in the open, named and pitched and waiting: I don't really want a bigger notebook. I want a collaborator. The point was never the ideas — it was finding the people to build the next ones with. Pick one. Tell me why it's wrong, or why it's right. The best things I've made started as a conversation.