On Monday evenings, Amir Hanna would come home from his engineering job and begin a second shift in front of a camera. He filmed about 20 short videos at a time, studying whatever had worked the previous week and changing one thing for the next. The arrangement had the practical romance of a garage startup, minus the garage and plus a full day at Shell. He was teaching himself online marketing in public, one awkward clip after another.
Hanna has said the first 600 were terrible. The number is funny until you imagine making them. He ultimately published 627 videos himself while still employed full time. His rule came from advice he associated with MrBeast and Shaan Puri: post every day for 100 days, improving one detail each time. The result was more than a content archive. It was a founder acquiring distribution through muscle memory.
Chapter one
The reluctant language-app founder
Parrot, the company Hanna now leads, turns Spanish practice into something deliberately familiar: a vertical feed of short videos from native speakers. Learners can follow interactive transcripts, tap and save words, repeat phrases and review vocabulary later. The phone still invites a swipe. The destination changes.
The idea did not arrive as a beam of founder certainty. Hanna's first response was doubt. When his cofounder Erik Dahl proposed another language app, Hanna saw the giant incumbent, the crowded app stores and a question with an unflattering answer: why should they add one more? That skepticism became useful because the team did not resolve it with a motivational speech. They investigated.
Hanna says the founders downloaded 52 language apps and used each for a few hours. They interviewed 65 learners across Reddit, Discord and their own circles. They found repeated affection for streaks and simplicity, alongside a recurring frustration: a person could accumulate points yet still struggle to understand a native speaker. The team built a modest first version, essentially a video player with captions, and took it to students. According to Hanna, 60 percent paid for early access. Doubt had met a receipt.
“What started as doubt turned into a mission: to connect people by breaking down language barriers.”Amir Hanna
There was personal mischief in the choice. Hanna describes himself as a C-plus French student who dreaded language classes. His complaint is not that learning should require no effort. It is that the effort should resemble the skill. To understand conversation, one must hear a great deal of language in context. Parrot organizes that exposure around comprehensible input, presenting material learners can mostly follow while stretching them with unfamiliar words.
Before the feed
An engineer notices wasted motion
Hanna's path to consumer education passed through mechanical engineering at the University of British Columbia. In 2020, he worked on Flora, a waste-collection optimization project that combined sensors with an analytics dashboard. Flora won the Smart Cities track of Microsoft's Discover AI challenge, placing first among 577 participants, and received a C$9,000 grant from the UBC student society's sustainability fund.
The subject changed from rubbish routes to language feeds, but the instinct is recognizable. Find motion that produces too little value. Add feedback. Re-route it. Parrot looks at the thumb's habitual journey down a screen and asks whether the same gesture can carry accent, meaning and cultural context.
After a stint with Harvest Builders, Hanna joined Shell's mechanical engineering team in Alberta. He announced the role with an unusually honest list of what the celebratory job post concealed: ghosted emails, rejections, interviews followed by more rejections and hours of preparation. “Keep pushing,” he wrote. Years later, that advice looks less like professional-network piety and more like an operating system.
Oliver Manilow, the founder behind that partnership, tells a story that makes the period vivid. In April 2025 he cold-called Hanna, who answered from Shell's Edmonton office. Manilow's company was mid-pivot and had little traction. Hanna had no marketing budget to offer. They spoke for 55 minutes anyway, two builders comparing notes before either had much leverage.
They stayed in touch. As Parrot grew, Manilow's team helped assemble a creator program, hiring 15 creators and, by his account, generating 45 million views and more than 10,000 customers before handing the system back to Hanna. The anecdote is a small corrective to the lonely-founder portrait. Distribution here was built from relationships: a call answered, ideas exchanged, trust compounded.
The inflection
When the night job became the job
Hanna left Shell in July 2025 to work on Parrot full time. By September, the company had joined Y Combinator's Fall 2025 batch and moved to San Francisco. Founded in 2024 by Hanna, Dahl and Julia Hudea, the three-person founding team met through a UBC entrepreneurship class.
Their roles make a tidy but human triangle. Dahl's language-learning obsession supplied the original provocation. Hudea brought a product and engineering perspective. Hanna, once the skeptic, became CEO and the team's most visible student of distribution. He has described the founders as people who had built projects that might have worked but did not matter enough to them. Parrot did.
The bootstrapped climb, as reported by Hanna
The numbers accelerated in Hanna's public telling. Parrot entered YC at $16,000 in monthly recurring revenue. Six weeks later, he reported $75,000. At eight and a half months from the subscription launch, he reported $79,000, with 7,300 customers and 3,000 active trials. The creator network had grown from eight people to 39. For Hanna, the conclusion was plain: as software becomes easier to make, distribution, brand and viral loops become more valuable.
Those metrics are snapshots from the founder, and early-company numbers have a habit of changing before the ink dries. The more durable observation is about product and behavior. Parrot does not demand that users become different people before they begin. It meets a familiar reflex and tries to improve its yield. There is behavioral judo in that, but also a moral question: can a productive feed preserve the pleasure of discovery without inheriting every vice of the infinite scroll?
Parrot's answer is progress that reaches beyond the app. In one 2026 update, Hanna wrote about a learner named Lucas who had clicked more than 18,000 Spanish words, saved 12,177, maintained a 208-day streak and logged 88 hours of listening. Hanna translated the final number into his preferred internet dialect: roughly 30 viewings of Interstellar. The joke carries the product's ambition. Screen time is going to exist. What might a person have to show for it?
“As software becomes easier to build, the real competitive edge shifts to distribution, viral loops, and brand.”Amir Hanna
The larger sentence
Language as a door, not a score
Hanna's personal biography supplies the emotional scale. He identifies as Egyptian-Canadian. His site says he speaks four languages and has visited 38 countries. In announcing Parrot's YC acceptance, he spoke about the practical power of language in Egypt: another language could open international work, but it could also add color to travel, depth to relationships and nuance to humor. Translation software can deliver a sentence. It cannot reproduce the warmth of choosing to speak to someone in their own language.
This is why Parrot's mission reaches beyond a more efficient lesson. The founders want to connect people by breaking down language barriers. That aspiration is charmingly large for an object operated with one thumb. It also keeps the feed accountable. Engagement is useful only if it ends in understanding.
Hanna's career is young enough that any finished portrait would be dishonest. Parrot remains a small company in a crowded market, with the ordinary hazards of consumer subscriptions, changing platforms and attention that moves capriciously. Yet his method is already legible. Begin skeptical. Test the field. Make the primitive version. Ask people to pay. Do the unscalable work yourself until you understand it well enough to build a system.
And keep a sense of proportion. A founder who has made 627 videos is allowed to laugh at the first 600. Most careers are edited to preserve only the clip that worked. Hanna's story is better with the outtakes left in. The imperfect repetitions explain the result and offer something another builder can actually borrow.
The scroll is usually treated as a weather system, vast and inevitable. Hanna treats it as an engineering constraint. The thumb will move. The feed will continue. Somewhere between one video and the next, perhaps a stranger's speech will become intelligible. That is a modest miracle disguised as product design.