Seven apps. All running simultaneously. All half-built. All "almost done." That was me six months ago, and if you're a solo founder with ADHD, you probably know exactly how that happens. Building 7 apps at once isn't a failure mode - it's the default state when your brain treats every new idea like an emergency and you're technically capable of actually building all of them. I can do the full stack. Design, frontend, backend, smart contracts. Which sounds like an asset until you realize it means nothing stops me from starting something new at 2am because I read one interesting tweet about a framework I've never tried. Let me tell you what that actually looks like. How building 7 apps at once happens (faster than you'd think) It doesn't start as 7. It starts as 1, then 2, then "just a quick experiment," then you blink and your Notion has 7 active project docs and your GitHub has 11 repos created in the last 90 days. My list at peak chaos: OpenClaw - my AI agent system. The main thing. 14 cron jobs running daily tasks, managing my workflow, aggregating data, handling research. Actually running in production. OpenClaw A music generation pipeline - GPT-4 generating lyrics, Suno turning them into actual audio, Midjourney for visuals, ElevenLabs for voiceover. The idea was a fully automated artist that produced content without me touching anything after setup. I got it working. Then got distracted. A music generation pipeline Crypto products - I've shipped tokens that hit $10M+ market cap. The financial incentive is real. The problem is crypto moves fast and demands constant attention, which ADHD brains are simultaneously great and terrible at. Crypto products An ADHD planning tool - I was going to build the productivity app that actually worked for my brain because everything else sucked. Classic "scratch your own itch" trap. You know what happened. I got distracted building it. An ADHD planning tool MCP servers - Model Context Protocol is genuinely interesting infrastructure work. I built three of them in a weekend because the architecture grabbed me. MCP servers CLI tools - Small stuff. Quick utilities. The kind of thing you build in 4 hours because you needed it for something else and now it exists. CLI tools Automation workflows - n8n pipelines, custom agents, glue code connecting APIs. This bleeds into everything else so it's simultaneously its own project and embedded in all the others. Automation workflows Seven things. One person. No cofounders by choice - I'll get to why in a minute. The ADHD founder loop is real and it's not what you think Everyone explains ADHD as "gets distracted easily." That's not wrong but it misses the mechanism. It's not that I can't focus. It's that my brain allocates attention based on novelty and interest, not importance and priority. A new framework registers as urgent. A 6-week-old project that needs documentation registers as optional. The loop looks like this: New idea hits. Immediate hyperfocus. Build for 16 hours straight.Get it working. MVP ships. Dopamine.Hit the boring middle - marketing, SEO, documentation, polish, customer support.Brain de-prioritizes it. New shiny thing appears.Repeat. New idea hits. Immediate hyperfocus. Build for 16 hours straight. Get it working. MVP ships. Dopamine. Hit the boring middle - marketing, SEO, documentation, polish, customer support. Brain de-prioritizes it. New shiny thing appears. Repeat. The boring middle is where non-ADHD founders grind through and ADHD founders bounce. The problem isn't the building. I'm fast at building. The problem is everything after the building. Here's what took me too long to figure out: the boring middle is where the money is. Shipping is fun. Distribution is how projects survive. And I can't do distribution manually. It just doesn't happen. This isn't unique to me - the ADHD and entrepreneurship research from Barkley et al. consistently shows that executive function deficits hit hardest on sustained, low-novelty tasks. Which is exactly what distribution and maintenance are. The building phase is fine. Everything after is structurally harder for ADHD brains. ADHD and entrepreneurship research from Barkley et al. What building 7 apps at once as a solo founder with ADHD actually costs you The real cost isn't time. It's context switching overhead compounded across seven projects simultaneously. Every time you context-switch between codebases, you pay a mental re-load cost. Research on developer context switching puts the refocus time at 23 minutes on average after an interruption. Multiply that by seven projects and you're paying that tax constantly. You're never fully in any one thing long enough to hit the deep work state where hard problems actually get solved. Research on developer context switching My peak chaos period felt productive. I was always building something. But my actual output - measured in shipped features, not hours of activity - was lower than my focused periods. The context switching was eating the work. The other cost is quality. When you're seven projects deep, you're not doing your best work on any of them. You're doing good-enough work on all of them. Which means none of them have the level of craft that makes a product actually stand out. I shipped an MVP of the ADHD planning tool that was functional but ugly. No time to polish it. Meanwhile I was debugging MCP server architecture and planning the music pipeline. Something had to give and it was always the boring middle tasks - the ones that determine whether anyone actually uses the thing. The solo founder reality nobody posts about how ADHD and AI work together is performance art now. Revenue screenshots get thousands of likes. MRR milestones. "I quit my job and 6 months later..." threads. Everyone's sharing the wins. how ADHD and AI work together Posts about almost going broke get 12 likes. Posts about abandoning your fourth project in a row get silence. The authentic failures don't get engagement so nobody posts them, which means every new solo founder thinks everyone else has it figured out. I've been broke. I've shipped things that made zero dollars. I've built for three weeks and then deleted the repo because I couldn't figure out what the product actually was. I've had projects that made real money and then lost relevance six months later because I didn't maintain them. The solo part: genuinely fast. No meetings about meetings. No "we should align on the roadmap" conversations with a cofounder who has different priorities. No equity negotiation. No founder breakup. I've watched every cofounder horror story play out in real time around me - equity fights, one person doing 80% of the work, fundamental disagreements about direction, personality conflicts that kill companies that should have made it. Solo is lonely. It's also fast. The ADHD + solo combination creates a specific failure mode though. There's no one to tell you you're going down a rabbit hole. No one to say "we've been working on this side feature for two weeks, should we refocus?" Your own judgment is the only check on your impulses and ADHD makes your judgment unreliable at the exact moments you need it most. The fix isn't a cofounder. The fix is building systems that act like a cofounder for the structural stuff while you handle the actual building. What actually survived (and why) Out of seven active projects, two are still running and one of those is making consistent money. Five are abandoned or on permanent ice. Here's the honest accounting: those five failures weren't wasted. They taught me things that made the two survivors better. I learned how to architect agent systems from the music pipeline experiment. The MCP servers gave me infrastructure patterns I'm using in OpenClaw. The ADHD planning tool taught me what actually helps my workflow versus what sounds good in theory. But the reason the two survivors survived isn't just that I learned from the failures. It's that the survivors automated their boring middles. OpenClaw runs 14 cron jobs. Morning digest, memory management, content scheduling, health data aggregation, crypto position monitoring - all automated. I talk to it every morning. The daily feedback loop exists. The daily maintenance doesn't. The crypto products that worked had simple mechanics and network effects that didn't require my constant attention. Once they were live, they ran. The ones that needed daily management died when my attention moved elsewhere. The pattern is consistent: if I have to manually do something for a project to stay alive, that project will eventually die. Not because I don't care about it. Because ADHD means I'll miss 3 days of manual tasks in a row and then it feels too far behind to recover and then I move on. If the system can sustain itself through a 2-week hyperfocus absence, it lives. If it can't, it doesn't. Why I'll never take a cofounder Everyone says this is a mistake. "Two founder teams outperform solo founders statistically." I've read the YC essays. Here's my actual take: those statistics are describing average cases and I'm not an average case. What a cofounder typically brings: accountability, complementary skills, emotional support during hard stretches, a second opinion on decisions, someone to handle the stuff you're bad at. What I actually need from that list: someone to handle the stuff I'm bad at. Specifically, distribution, SEO, documentation, customer communication. The boring middles. A human cofounder would need 40-50% equity for that. They'd have opinions about the product direction. We'd need to agree on decisions. We'd have to coordinate schedules. They'd have good months and bad months and burnout cycles that don't sync with mine. An AI agent running my content pipeline costs me $40/month and doesn't have feelings about feature prioritization. I'd rather have a multi-agent system that handles my weak spots than a human who wants half the company to do the same job. That's not cynical, it's just math. And it only works because I can build the agents myself. The loneliness is real. I'm not dismissing it. Some weeks the silence is rough. But the alternative isn't "cofounder" - it's growing on Twitter and genuinely engaging with the community instead of just posting screenshots. growing on Twitter The systems that make solo + ADHD actually work I've been iterating on this for two years. Here's what actually helps: File-driven everything. My agents read from markdown files - SOUL.md for personality, AGENTS.md for protocol, MEMORY.md for context. When something needs updating, I edit a text file. No deployments. No configuration hell. Low friction maintenance survives ADHD gaps. High friction maintenance doesn't. File-driven everything. Cron jobs over manual triggers. If a task requires me to remember to do it, it will eventually not get done. Scheduled jobs run whether I'm in a hyperfocus sprint or a three-day crash. Everything recurring gets automated. Cron jobs over manual triggers. Ruthless scope decisions at the start. The ADHD planning tool died partly because I kept adding features during the build phase. Every new feature felt urgent. Scope compound-expanded until the MVP was a six-month project. Now I write down the exact features for v1 and I'm not allowed to add anything until v1 ships. This one rule has probably doubled my shipping rate. Ruthless scope decisions at the start. Letting projects die without guilt. I spent too long trying to "get back to" projects I'd mentally left. Felt like failure. Now I have a graveyard folder on GitHub and I use it. Dead projects get documented (so I can steal the code later) and archived. The folder has 22 repos in it. That's fine. Letting projects die without guilt. Using the hyperfocus instead of fighting it. When I'm locked in on something, that's a resource. I try to batch the hardest technical work for those windows. The structured boring tasks - docs, config, cleanup - get batched for the lower-energy periods when I can work mechanically. Using the hyperfocus instead of fighting it. Finishing 2 out of 7 is actually fine Here's the reframe that helped me most: the 5 abandoned projects aren't failures, they're the cost of finding the 2 that worked. Most indie developers don't talk about their abandonment rate. They talk about the thing that shipped. But every shipped product has a graveyard of experiments behind it. The experiments weren't wasted - they were the research. Most indie developers My music generation pipeline taught me how to build multi-tool AI pipelines. Spent two weeks on it, never shipped it as a product, use those patterns constantly now. The ADHD planning tool taught me that "I'll use this myself" is not the same as "other people will pay for this." Different lesson. Equally valuable. The CLI tools I built in weekends are still sitting in my shell. I use them every day. Zero revenue. Huge quality-of-life improvement. Not everything needs to be a business. The crypto tokens that hit $10M+ market cap came after multiple experiments that went nowhere. The ones that worked worked because of things I learned from the ones that didn't. Building 7 apps at once as a solo founder with ADHD is chaotic. It's also how you discover which two are actually worth finishing. If you're building solo with ADHD and you've got four half-built things on your GitHub and you feel like you're failing - you're not. You're doing the process. The trick is setting up the systems so that when the thing that matters is ready to run, it can run without you holding it up. Automate the boring middle. Let the losers die fast. Stay technical enough to build your own infrastructure. And ignore anyone who tells you the only way to do this right involves giving half your company to someone else. The agent handling my distribution doesn't want equity. That's the whole point. If you want to see how the agent system I built actually works - the one that survived while six other projects didn't - I write about it here. The full architecture, the cost breakdown, what broke, what didn't. agent system

GitHub

Learning Finance by Designing It

Portfolio

Building 7 Apps at Once With ADHD Nearly Broke My Solo Founder Brain

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

Accessible Data Visualization in Fintech: Why It Matters

The Noonification: Use This 7-Step McKinsey Framework to Solve Any Problem (1/10/2023)

The Noonification: A Taxonomy of Inclusiveness (1/11/2024)

The Noonification: What is the InfiniteNature-Zero AI Model? (11/19/2022)

10 Ways AI Has Changed Our Lives

100 Days of AI, Day 8: Experimenting With Microsoft's Semantic Kernel Using GPT-4

Accessible Data Visualization in Fintech: Why It Matters

The Noonification: Use This 7-Step McKinsey Framework to Solve Any Problem (1/10/2023)

The Noonification: A Taxonomy of Inclusiveness (1/11/2024)

The Noonification: What is the InfiniteNature-Zero AI Model? (11/19/2022)

10 Ways AI Has Changed Our Lives

100 Days of AI, Day 8: Experimenting With Microsoft's Semantic Kernel Using GPT-4

Light-Mode

Classic

Newspaper

Minty

Dark-Mode

Neon Noir

Minty

HN StartUps