Before it’s everywhere, it’s in FWDstart. Subscribe for $99 and save 17% on the first year to stay ahead.
Aly Moursy has spent much of this year embodying an interesting contradiction of the AI era.
Which is to say that, over the past several months, he's been trying to automate as many of his company's functions as possible while simultaneously making himself as visible and accessible as he can to prospective customers.
To my mind, it's one of the better illustrations I've come across of what an AI-native startup might actually look like, a term that's been bandied about with considerable abandon and applied to just about every nascent company whose founders have a Claude Max subscription.
Such is the impact agents have had on the business of building software over the past several months that describing 2026 as the year of the agent would somehow still feel like an understatement.
If you're on the hunt for a good explanation as to why all of this has suddenly become possible, OpenAI CEO Sam Altman offered one in a conversation with Stripe co-founder John Collison last week, explaining how advances in reasoning and inference-time compute have made it possible to throw entire teams of agents at increasingly complicated problems, rather than relying on a single model to work through everything itself.
It's opened up something of a chasm between what most companies are doing with AI and what its most sophisticated users are already accomplishing.
🔓 The full version of this article is for paid subscribers only.
Moursy might be one of the more colourful examples I've encountered of what the latter actually looks like.
The Egyptian entrepreneur, whose Y Combinator-backed startup Veeza AI is attempting to automate much of the bureaucratic tedium and misery involved in applying for tourist visas, has assembled an increasingly elaborate and dizzying collection of coding agents and personal assistants that work concurrently, coordinate with one another and perform functions that, until fairly recently, might have required an entire team of people.
He's even reached the point of building agents whose job is to improve the agents he's already created.
"This is now the next level of my LLM psychosis," he joked during our conversation.
What's perhaps more interesting, though, is that the further he's pushed automation, the more convinced he's become of the value of putting a human face to the business.
If you haven't already come across Moursy's memes, videos, demos, walkthroughs or increasingly elaborate experiments with AI somewhere on social media, I'd be surprised, such has been the reach of what has proved to be a rather compelling distribution strategy.
So much so that Veeza AI's Y Combinator launch generated more engagement on LinkedIn than any of the other 54 Fall 2026 launches tracked by Fiber AI.
"When the marginal cost of software is going to zero," Moursy told me, "you are the only way people are going to understand that A is legit and B isn't."
His advice to fellow founders is straightforward.
"You need to be on camera."
The human distribution engine
For most of his life, Moursy…
Become a paid subscriber today to continue reading
To support our journalism, and unlock all of our stories and original analysis about the the MENA startup and VC ecosystem, subscribe below.
