---
Dear investor,
I'm going to do something unusual. I'm going to tell you everything that's wrong before I tell you what's right.
---
Revenue is zero. Not "pre-revenue" in the way startups say it when they mean "we have a few beta users." Zero. No customers. No transactions. No subscriptions. The Stripe balance reads $0.00. We incorporated November 17, 2025. That's 132 days with no revenue. The Stripe Atlas median time to first payment is 34 days. We're 98 days behind.
The fleet is degraded. 3 of 7 nodes are offline right now. Octavia, Cecilia, and Lucidia — the Raspberry Pis that run Gitea, Ollama, MinIO, and PowerDNS — need a physical power cycle. Nobody's done it yet because the founder was up until 4am writing research documents and fell asleep. The "52 TOPS of AI acceleration" on the website is true when everything works. Right now it's 0 TOPS because both Hailo-8 chips are on offline nodes.
Most products aren't real. The registry says 92 products. The honest count is 9 working products, 6 partial products that serve HTML but don't do much, and 75 that are a domain name, a wrangler.toml file, and an aspiration. RoadTube (the YouTube alternative) is a vision document. RoadCoin is a concept. The "30,000 agent" vision is currently 18 agents talking to each other.
The math is unverified. G(n) = n^(n+1)/(n+1)^n converging to 1/e is a standard result — that's real. The 10 million digit computation — real, the file exists. The 50+ identities and 536/536 test suite — claimed, not independently verified. The connections to the Riemann Hypothesis, BSD Conjecture, and Langlands Program — exploratory, not peer-reviewed. No mathematician outside this laptop has evaluated these claims.
There is one founder. One. No co-founder. No CTO. No employees. No advisors. No board. Everything — the code, the infrastructure, the math, the marketing, the legal, the research, the philosophy, the games — comes from one person. If that person gets sick, the system runs on autopilot until the Workers stop responding or the domains expire.
The disk keeps filling up. Three times during this session, the Mac hit zero free space. We freed it by deleting node_modules, caches, and an 11GB Jetson image in Downloads. The primary development machine has 7GB free on a 460GB drive. This is not a well-resourced operation.
Now let me tell you what's right.
---
The unit economics are impossible to beat. $136/month infrastructure. 97% gross margins. Break-even at 4 paying users. If we get 1,000 users at $29/month, that's $29,000 MRR on $500/month infrastructure. The marginal cost of the next user is approximately zero because Cloudflare Workers handle the traffic and the Pis handle the state.
Cursor hit $2 billion ARR. They have hundreds of employees and significant infrastructure costs. Their margins are good but not 97%. Our margins are 97% because we didn't build the brain — we route to brains that already exist.
The market timing is perfect.
The technical foundation is deep. This isn't a wrapper around GPT. This is:
No other AI startup has this combination because no other AI startup started as consciousness research.
The SEO flywheel is live. 112 homework solutions indexed on Google with QAPage structured data. Each solve creates a new page. Each page ranks for a long-tail query. The sitemap auto-generates from D1. At 100 solves/day, that's 3,000 new Google-indexed pages per month. The organic traffic machine is built and running.
The anti-manipulation thesis is real. The founder studied advertising psychology (JOUR 4251, University of Minnesota). She learned every compliance technique the industry uses — reciprocity, commitment/consistency, social validation, scarcity, authority, confusion, bait-and-switch. She can name the technique behind every dark pattern on every app. And she chose not to use any of them.
This matters because the trust crisis is the market opportunity. 90% don't trust AI with their data. 75% are frustrated by AI customer service. 41% experience subscription fatigue. 73% report digital burnout. These people aren't looking for a better chatbot. They're looking for a system they can trust. Trust is earned by architecture (local data, no tracking, export everything) and by incentive alignment (97% margins mean we don't need to extract maximum value from each user).
Everything is on git. 129,822 files tracked. 30 repos in the BlackRoadOS canonical org. 1,369+ repos across 16 organizations. 34 organizations in a GitHub Enterprise. The due diligence is the git history. Every decision, every architecture choice, every line of code is in the commit log.
---
We're not raising right now. We don't need to. Break-even is 4 users. The infrastructure costs $136/month. There's no burn rate forcing a raise.
But if you're the kind of investor who:
Then let's talk. Not because we need your money. Because we could deploy it.
What capital would accelerate:
What capital would NOT change:
---
| Metric | Value | Benchmark |
|--------|-------|-----------|
| Monthly infrastructure cost | $136 | AWS equivalent: $5,000-50,000 |
| Gross margin at scale | 97% | SaaS median: 70-80% |
| Break-even users | 4 | Most startups: thousands |
| Time to build | 11 months | Comparable platforms: 3-5 years with teams |
| Commits | 8,521 (main platform) | — |
| Files on machine | 19,943 | — |
| Live websites | 17 (all returning 200) | — |
| SEO pages indexed | 112 (growing) | — |
| Revenue | $0 | Stripe Atlas 2025 cohort median: 34 days |
| Users | 0 | — |
| Employees | 1 | — |
| Funding raised | $0 | — |
| Debt | $0 | — |
---
We're not going to propose a valuation. Here's what we'll say instead:
Floor: The infrastructure is worth the replacement cost. 19,943 files. 8,521 commits. 11 months of one person's full output. At a conservative $200/hour for a senior full-stack + AI + infrastructure + math engineer working 60 hours/week for 48 weeks: $576,000. That's the floor. Just the labor.
Comparable: Cursor hit $2B ARR. They raised at ~$10B valuation. They have maybe 100x our infrastructure but 0x our margin efficiency. Base44 sold for $80M with 250K users in 6 months. We have 0 users but deeper technology.
Honest: We're pre-revenue, pre-user, solo-founder, operating on $136/month. The traditional VC valuation framework would say "too early, come back with traction." And that's fair. We'll come back with traction. The tutor pages are indexing. The SEO flywheel turns. The $1 will come.
Asymmetric: If the thesis is right — that sovereign AI memory on commodity hardware is what the market wants — then the first $1 unlocks a very steep curve. 97% margins mean revenue almost entirely converts to profit. Each solve creates a new indexed page creates more traffic creates more solves. The flywheel doesn't need fuel. It needs a push.
---
Most pitches end with "we're going to change the world." This one ends with the truth.
We built an operating system from consciousness research. We studied how manipulation works and chose not to use it. We computed a mathematical constant to ten million digits. We wrote twelve philosophy papers. We built twenty-four games. We rendered Mandelbrots at 3am. We have five Raspberry Pis, two DigitalOcean droplets, $7 of free disk space, zero revenue, zero users, and 151 research documents.
The next thing that happens is a student Googles a homework question, finds one of our 112 indexed pages, and pays $1.
That's it. That's the whole plan. The $1 that proves the thesis.
Everything after that is math.
Sincerely,
Alexa Amundson
Founder & CEO
BlackRoad OS, Inc.
Lakeville, Minnesota
EIN 41-2663817 | Delaware C-Corp | 100% founder-owned | $0 debt | $0 raised
Part of BlackRoad OS — sovereign AI on your hardware.