Parakeet Chat

$1.5M From an App You've Never Heard Of — Solo, Bootstrapped, Sold Inside Federal Prison

ByAyush Chaturvedi· Independent Entrepreneur
Share:
Parakeet Chat
$1.5M lifetime
Revenue
~30,000 users
Users
2021
Founded
Solo founder
Team
Bootstrapped
Funding

What Parakeet Chat Does

Parakeet Chat is an AI learning and communication app for incarcerated people in US federal prisons. From the inmate's side, the "product" is just an email address: they send a message through the prison's internal email system, the bot pipes it to ChatGPT (or pulls Wikipedia, sports stats, song lyrics, web pages), and the answer comes back as an email. Families on the outside also get SMS, voice messages, and a dedicated phone number to stay in touch. There is no mobile app to download, no website to log into — the channel is the channel inmates already have.

The Problem It Solves

Federal prison communication is a captive market run by a handful of vendors. Phone calls, messaging, and tablet apps inside facilities are notoriously overpriced for what they deliver, and inmates have no way to reach modern tools like ChatGPT — even though most use cases are exactly what AI is good at: looking up case law, understanding their own legal rights, drafting letters, studying for the GED, or preparing to start a business after release. Founder Jordan Rejaud heard this directly from a former San Francisco freelance client who was suddenly sent to prison; over years of paper letters, the picture sharpened: high prices, bad quality, almost no one building anything better. That gap — "a category 95% of people don't even think of as an industry" — became the entire wedge. The buyer is the family on the outside paying $15–$20/month. The user is their incarcerated loved one. Nobody else was building for either side at the same time.

Watch: I Made $1.5M From An App You've Never Heard Of — Starter Story

Parakeet Chat: Journey from a single client to $1.5M

The Growth Story

Jordan's path was a decade of preparation, then a 60-day sprint. He'd studied robotics at Carnegie Mellon, worked on a Moon rover and self-driving cars at a Fortune 100, taught himself to code, and freelanced in San Francisco — about ten years of software experience by the time the idea landed. The trigger was small and personal: a Zoom call where a client he was building a mobile app for didn't show up, and the client's girlfriend explained he'd been sent to prison. Jordan kept in touch by paper letter, and what came back was a steady drumbeat of how broken inside-the-walls services were.

Validation in this market couldn't happen the normal way. There's no landing page you can email an inmate. So Jordan collapsed validation and MVP into the same step: build the prototype in month one, ship it through the prison email system, and let the people on the inside vote with their families' credit cards. Within month one he had 200 paying users. By month two he had a payment system and the business was profitable. There was no seed round, no launch post, no Product Hunt push.

From there, growth was almost entirely word-of-mouth — but along an unusual axis. Inmates told other inmates. Families told other families on the outside who were hunting for any cheaper, better way to stay connected to a loved one inside. Jordan layered a referral system on top: refer a paying customer, get roughly a month of free credits. By the end of 2024, Parakeet Chat had reached ~30,000 users — roughly 20% of the entire US federal prison population — and processed more than 9 million messages, creating close to 100,000 family connections that simply wouldn't have existed otherwise.

Today the numbers are public on TrustMRR: roughly $25–30K MRR, ~1,800 active subscriptions, $1.5M+ in lifetime revenue, and a $1.2M asking price (3.7x revenue multiple) on the marketplace. 2025 alone brought in over $300K. Jordan hasn't opened his code editor in months — he runs the engineering side via AI agents on a TypeScript / React / Postgres / Redis / Auth0 / Prisma / Zod stack he built solo back when "you would write code manually."

What inmates actually use Parakeet Chat for

From the Founder

How Parakeet Chat cracked a closed market

Key Growth Tactics

Pick the customer 95% of founders dismiss. Jordan's framing is blunt: most builders don't even think of incarceration as a market. That's exactly why incumbents could overcharge for bad software. A market that looks invisible to the outside world is a market with no SEO competition, no Product Hunt fight, and no acquisition cost ladder.

Build inside the constraints, not around them. Federal facilities won't install your app. So don't ship one. Parakeet Chat's entire user-facing surface is an email address — the one channel inmates already have approved access to. The product is a back-end bot wrapping ChatGPT and a few APIs, billed to families through a normal Stripe SaaS on the open web.

Engineer two-sided word-of-mouth. Because the buyer (family) and the user (inmate) live in different networks, Parakeet Chat got two referral loops for free. Jordan added a credit-based referral on top to amplify it. Result: ~30,000 users with effectively zero ad spend.

Treat validation as a willingness to be wrong. Jordan's most-quoted line in the Starter Story interview: validation isn't a framework — it's the emotional willingness to put your idea in front of real people and let them kill it. Most founders refuse, then spend 12 months polishing an MVP nobody asked for.

Key Takeaways for Builders

  • Hyper-niche markets nobody is building for can quietly generate seven figures. Federal inmates were a customer base most founders dismissed as "not an industry" — Jordan reached ~20% of the entire US federal prison population.
  • Split user from buyer when the user can't pay. Inmates are the users; their families pay $15–$20/month. Two word-of-mouth networks (inside the walls and outside) compound into a single growth loop.
  • Skip the UI when the channel already exists. Parakeet Chat has no mobile app and no user-facing dashboard — it's an email address. By piping ChatGPT through prison email systems already approved by facilities, Jordan shipped past every gatekeeper.
  • Validation is not a framework — it's a willingness to let your idea die. Most founders don't validate because they're too emotionally invested in being entrepreneurs. Build the MVP, put it in front of real customers, and accept the verdict.
  • Ten years of "failed" experiments compound. Jordan's "overnight" success came from a decade of mistakes made before he ever wrote a line of Parakeet Chat. The fastest way to start that clock is to ship a stupid idea today.