On February 2, 2023, a fourth-year computer science student in Waterloo tweeted a 30-second demo video to his 16 followers.
That tweet is now the first line of a $10 million revenue story.
The student was Yasser Elsaid. The product was Chatbase — then a crude "ChatGPT for your PDFs" tool, today an AI agent platform for customer experience used by more than 10,000 businesses. One hundred and seventeen days after that first tweet, Chatbase crossed $1 million in annualized revenue. Zero venture capital. Zero ad spend. A marketing budget of, literally, nothing.
Every founder on earth claims their MVP story is different. This one actually is — because every number in it is public, founder-confirmed, and independently verified. So let's take it apart properly: what Yasser did right, what he got lucky on, and what you should copy the next time you're staring at an empty repo at midnight.
Here is the MRR ramp, in Yasser's own words from his Indie Hackers interview with Channing Allen:
| Date | MRR |
|---|---|
| Feb 7, 2023 | $0 |
| Feb 11, 2023 | $400 |
| Feb 16, 2023 | $900 |
| Feb 28, 2023 | $3,000 |
| Mar 15, 2023 | $10,000 |
| May 13, 2023 | $64,000 |
May 13 is day 101 of the year. $64K MRR is a $768K annual run rate — and per Superframeworks' deep dive, Chatbase crossed the $1M annualized mark exactly 117 days after the launch tweet.
Put that in context. Startup Founder Stories benchmarks Yasser's journey against documented founder journeys: he hit his first paying customer in 7 days (median: 1 month), $10K MRR in 45 days (faster than 93% of 295 documented $10K MRR journeys; the median is 12 months), and $100K ARR in 4 months (median: 2 years — he was 84% faster).
And the story didn't peak in 2023. December 2025: $8M ARR, 92% retention, 18 people. May 2026: $10M ARR, 10,000+ customers — a figure Stripe independently confirms, and a trajectory GetLatka tracks with no outside funding. Profitable the whole way.
The origin story matters because it's the opposite of a master plan.
Yasser moved from Egypt to Canada in 2019 for undergrad. He did everything right by the traditional script — landed internships at Tesla and then Meta. And he hated both. His words: the lifestyle was "too safe, too manicured." Same day, every day. Those internships taught him he wanted to be an indie hacker.
So he swapped leetcode grinding for side projects. Most went nowhere. Then in his fourth year he started playing with the OpenAI API — before ChatGPT made AI mainstream — and spotted the gap himself: "powerful models, no access to your specific data."
His insight into his own success is worth quoting because it's honest: "I was at the right place at the right time... Also, my tweets going viral definitely helped. Those are forces that are outside of my control that contributed to Chatbase's success. But to give myself some credit, I also worked really hard (12-hour days) to capitalize on the opportunity."
Keep that duality in mind. Luck and positioning set the table; he still had to eat.
Here's what the actual MVP was: upload a PDF, chat with its contents. That's it. Two months of building — one person, no co-founder, no team.
And here's what he skipped, and this is the part that should break some brains: zero validation, zero market research. His own words: "I didn't do any validation or market research since this space was just being born and it was obvious to me that 'ChatGPT for your data' would become a thing."
Before you screenshot that as permission to skip discovery forever — look at what he actually did instead. He validated through velocity: he was already experimenting with the OpenAI API months before launch, and when the tweet showed demand, he went 100% all-in the same day. The validation wasn't skipped. It was compressed into days instead of quarters.
The stack was equally unromantic: React, Next.js, Supabase, OpenAI API, Langchain, Pinecone, Stripe. Boring, proven, well-documented. Nothing custom until money forced it.
And the pricing discipline most first-timers fumble: he launched with a pricing page — not a free beta, not a waitlist — and the first Stripe notification arrived 30 minutes later. The second, 10 minutes after that. Most revenue eventually came from the $399/month plan. A solo founder with a free tool is a charity. A solo founder with a $399 plan is a business.
The demo itself was engineered to travel: a screen recording of a chatbot trained on a document that explained how Chatbase worked. The demo demonstrated the product, using the product. Self-referential, instantly graspable, with the "wow" factor of chatting with your own files — riding the crest of ChatGPT hype.
It went viral. AI influencers like Rowan Cheung posted about it unprompted. Then Yasser worked the follow-on channels deliberately: a Product Hunt launch, story-framed posts in relevant subreddits ("a story works better than 'here is a tool I made'"), activity on Indie Hackers, submissions to AI directories.
His first two hires — a friend for engineering and someone for customer support — came straight from his own bottleneck: he was answering support emails from his personal inbox. Hire for the pain you feel right now, not the org chart you imagine in three years.
Now the part most MVP retrospectives leave out, because it's uncomfortable: the thing that blew up was also the thing that couldn't last.
"ChatGPT for your PDFs" was the easiest use case to copy. Within months, dozens of near-identical tools flooded the space, and being first stopped mattering. Yasser saw it clearly: virality "is not a good long-term strategy," and a single OpenAI release decision "will change the trajectory of all these AI tools."
So Chatbase repositioned — from a consumer-ish novelty to a platform for customer-facing AI agents that take actions in B2B environments. Harder to build, much larger budgets. Today chatbase.co sells agents across chat, email, and voice, with procedures, guardrails, a built-in helpdesk, SOC 2 Type II, GDPR and HIPAA compliance, and customers like Chuck E. Cheese and Les Mills. OpenAI's own Head of Startups, Marc Manara, is quoted on the site calling Chatbase "a strong signal of how customer support will evolve."
The team even moved from Toronto to New York using what Yasser calls the 98-out-of-100 rule: when they listed 100 ideal customer companies, 98 were in New York. So the whole team moved. Proximity to buyers beats proximity to comfort.
1. He surfed a wave he'd been studying before it crested. He was building with the OpenAI API in 2022. When ChatGPT detonated in November 2022, he wasn't learning the tooling — he was already shipping. You can't schedule luck, but you can pre-position for it.
2. He made the demo do the selling. Self-demonstrating product, wow moment in seconds, zero jargon. If your MVP needs three paragraphs of explanation, the MVP is the problem.
3. He charged from day 30 minutes. A pricing page at launch, first Stripe ping within the half hour. Revenue as the validation metric — not likes, not signups.
4. He priced for the business buyer. $399/month plans doing the heavy lifting meant a few hundred customers could equal millions in ARR. Low-ticket plans mean you need thousands of customers to matter.
5. He distributed through stories, not ads. Reddit stories, build-in-public tweets, community presence, AI directories — organic channels that compound. "No marketing budget. He didn't have one."
6. He shipped visibly to kill churn. By ProductLed's account, Chatbase leaned on constant visible shipping — "no cancellation trick beats a product customers watch improve." 92% retention by 2025 says it worked.
7. He rebuilt the moat before the wave receded. From commodity PDF chat to agentic CX platform. He cannibalized his own viral product on purpose, before someone else did it for him.
One caution before you sprint off: 2023's ChatGPT wave was a once-in-a-decade demand shock, and even Yasser calls the timing mostly luck. You can't re-run it. What transfers is the operating system — velocity over polish, revenue over applause, compounding distribution, and the discipline to kill your own golden goose before the market does it for you.
The MVP isn't the finish line. It's the first tweet. Ship it.
