Ahmed Omar learned his first lesson in internet commerce with a motorcycle and a request from his father. He was 14. The machine needed a buyer; the family needed the money. Omar put it online and sold it for AED 500, about $136 at the time. The amount was modest. The revelation was not. A person could take a stubborn, physical object, translate it into a listing and make a market appear through a screen.
That transaction has the tidy shape of a founder origin story, but the years that followed were full of messier practice. Omar sold through regional marketplaces while studying business in Cairo. He built Seyanty, a car-maintenance booking product, before he knew the vocabulary of venture capital. He tried Find My Pic, an app that let people retrieve photos with keywords, then watched its premise weaken when Google Photos arrived with a similar capability. The lesson was delivered without ceremony: a clever feature can be swallowed by a platform.
He also fought. Omar’s public record lists three MMA gold medals in Egypt, including titles in 2011 and 2012. Fighting and selling reward different instincts, but both produce an unsentimental scoreboard. The product moved or it did not. The judges raised your hand or someone else’s. This is the useful thread through Omar’s career: he likes measures that force activity to become an outcome.
01 · The first arena
A billboard-sized appetite
In 2014, after graduating from the German University in Cairo, Omar arrived in New York on a visitor visa. He has written about standing in Times Square, looking up at the billboards and deciding that one day his company would appear there. It eventually did. The anecdote lands because it is both grand and slightly playful: a new arrival using that famously loud advertising surface as a private vision board.
The next serious company was Odiggo. The problem belonged to a market that resisted neat software abstractions. Car owners across Egypt and the wider Middle East struggled to locate reliable parts and repair providers. Vendors were fragmented. Prices and quality varied. Omar and his collaborators built a marketplace to connect the pieces. In a 2018 interview, he said Odiggo had found 2,000 active customers in seven months. By 2021, the company said it served tens of thousands of users across three markets and worked with hundreds of merchants.
Odiggo entered Y Combinator’s Summer 2021 batch and announced a $2.2 million seed round with backing that included Y Combinator, 500 Startups and Plug and Play. It was a validating chapter, but not the final destination. The durable education was in stitching together a transaction across an industry whose information lived in many places and whose work happened offline. A marketplace succeeds only when the handoffs succeed.
02 · Six months before code
The workflow is the product
Before Sully.ai wrote code, Omar says he spent six months working as a healthcare administrator. That detail explains more than any model diagram. From the inside, the work did not present itself as a single sparkling AI use case. It arrived as a chain: intake, scheduling, chart review, the visit, documentation, coding, billing and follow-up. Each link had its own software and each handoff could return work to a human queue.
The obvious product in 2023 was an AI scribe. It could listen to a conversation and draft a note, an immediate demonstration with an immediate benefit. Omar and his co-founders chose a wider frame. Sully.ai would assemble a roster of agents with recognizable jobs: receptionist, triage nurse, scribe, medical coder, medical consultant. The names were commercial translation. A hospital buyer did not need to understand an orchestration layer to understand a team.
This was not merely a bigger bundle. It was a claim about where value hides. Traditional software records, displays and routes. Omar wants the system to execute the task and pass the result onward. In his preferred metaphor, analytics are the brain that notices 50 denied claims; the agent is the pair of hands that prepares the 50 appeals. The language is deliberately concrete because the buyer’s problem is concrete: something remains undone at the end of the day.
03 · A new scoreboard
Count the minutes that disappear
One revealing Omar idea is a metric. Many software businesses count daily active users, seats or sessions. Sully.ai uses Minutes Added to Workforce, shortened to MAW. It estimates the labor-equivalent time performed by its agents: call duration for a receptionist, conversation audio for a scribe, task-based time for generated coding work. The company reported more than 20 million such minutes during 2025.
The number is company-defined and company-reported, so it is not a neutral accounting standard. It is still a revealing choice. Daily activity treats attention as success. MAW treats the absence of required attention as success. A doctor who never opens the product because the work is already complete may be the ideal user. This flips the familiar SaaS dashboard inside out.
Conceptual illustration, not a performance comparison. The distinction is between measuring use and measuring finished work.
The metric also clarifies the company’s pricing story. Sully.ai has described customers as paying an annual “salary” per physician served, borrowing the language of payroll rather than software licensing. This packaging carries risk as well as force. Calling an agent an employee raises expectations about reliability, supervision and accountability. The metaphor wins attention, then creates a high bar the product must clear every day.
04 · Let the co-founder be right
The feature that escaped the backlog
One of Omar’s most useful product stories begins with disagreement. Ahmed Nasser, his longtime collaborator and Sully.ai co-founder, proposed an after-visit assistant that could retrieve information from anywhere, including a provider’s electronic medical record. Omar saw a good feature but wanted to deprioritize it. They released an initial version anyway. Users pulled it forward.
Omar’s conclusion was not the usual sermon about always listening to customers. It was about colleagues: do not stand in the way of exceptional people, even when they may fail, because a well-run failure generates learning. The idea has the texture of a founder who knows that decisiveness can curdle into obstruction. In a fast company, judgment includes knowing when to let someone else place the bet.
His earlier career supplied the less flattering version of that lesson. Looking back at Odiggo, Omar named two major mistakes: not hiring leadership early enough and not listening to customers from day one. Public founder stories often turn every bruise into destiny. Omar’s admissions are more practical. Growth exposes the work a founder has failed to delegate. A product exposes the assumptions its maker failed to test.
05 · Capital meets the metaphor
From a tool to a workforce layer
The market rewarded the framing. Sully.ai said it reached seven-figure contracted annual recurring revenue and more than 100 healthcare organizations within ten months of launch. In January 2025, it closed a $22 million Series A led by Amity Ventures, with participation from Phaze Ventures, Y Combinator, SemperVirens Venture Capital and other investors. Later that year, Omar reported more than 400 organizations and a team still early in what he calls a 30-year mission.
Those milestones do not settle the difficult questions. Hospital work crosses old systems, strict permissions and professional judgment. An agent that completes a task needs monitoring, defined boundaries and a clean path to a human. Coordination compounds usefulness, but it can also compound mistakes. The company’s task is to make the employee metaphor feel reassuring after the demo, not only memorable during it.
Omar’s public writing leans into the far horizon: autonomous systems that anticipate work, specialized agents that collaborate and far more care capacity than the current workforce can supply. Yet the credible version of that future will be earned in ordinary moments. Did the appointment get booked? Did the note arrive in the right field? Did the code move through review? Did the follow-up close? Grand missions survive on small completions.
06 · The fighter’s clock
Thirty years, measured one minute at a time
There is a productive tension in Omar’s style. He writes in large numbers and long horizons, but he organizes the company around minutes. He imagines a transformed global system, then returns to an appeal letter, a phone call or a chart. The former fighter still seems drawn to the round clock: aspiration is useful, but only the work completed before the bell changes the score.
His path from Cairo to New York, Dubai and the Bay Area can look improbably cinematic. The more instructive version is repetitive. List an object. Find a customer. Enter a fragmented industry. Connect the steps. Discover that a platform has erased your feature. Listen later than you should. Let a co-founder ship the idea you doubted. Change the company when the opportunity changes. Repeat.
Sully.ai is now a test of whether those habits can survive a domain where the stakes and the integration burden are unusually high. Omar is not selling the pleasure of using software. He is selling the relief of work no longer waiting. That distinction is easy to say and hard to build. It is also why his metric matters. A seat can be purchased and forgotten. A returned minute belongs to someone again.