We value your privacy.
Airtrek AI uses essential cookies for the app to work and optional analytics cookies (Google Analytics) to understand how people use the site. We do not sell your data. Read more in our Privacy Policy.
How Airtrek works: AI characters with era-locked knowledge, distinct voices, selfie-to-era portraits, persistent memory, and honest handling of uncertainty.
Airtrek is an immersive AI history experience. This page explains, honestly and in plain language, how the pieces fit together, the conversations, the voices, the portraits, the memory, and the limitations. If you want to know what the experience is before you try it, this is the page.
You pick a location and time period, Imperial Rome under Trajan, Ancient Greece in the age of Homeric song, Edo Japan under the Tokugawa shogunate, Gilded Age New York, or Ancient Egypt under Ramesses II. You arrive in a 3D rendered version of that world: the Roman Forum, the agora of an Ionian city, the harbor of 1880s Manhattan. You can look around the scene and navigate between views, the harbor, the hills, the great hall.
The scene is a backdrop, not a sandbox. It sets the place. The people you meet are what make it feel inhabited.
You can upload or take a selfie, and Airtrek transforms it into a historically grounded portrait of you as a person of that time and place, a Roman citizen, a Greek of the heroic age, an Edo townsman, a Gilded Age New Yorker. The portrait is generated by an image model and styled to the era's clothing, setting, and visual language.
This is not a filter. It's a way of seeing yourself in the scene, of making the past feel like a place you arrived in rather than a page you're reading. Your raw selfie is stored encrypted on your device, not uploaded as plaintext.
The core of Airtrek is conversation. You talk with characters who live in the era you've chosen, a senator, a merchant, a priestess, a poet, a queen. Each one has:
The characters are not one all-knowing narrator. They are multiple people with different perspectives, and the point is to talk to several of them and notice where they disagree.
This is the most important design constraint in Airtrek, and the one most AI history tools get wrong: each character knows only their own time and place.
A Roman senator in 110 CE does not know about the fall of the empire, it hasn't happened yet. Homer, singing in 8th-century BCE Ionia, does not know about Classical Athens, because it's three centuries in his future. A samurai in 1800 Edo does not know about the Meiji Restoration. A Gilded Age magnate in 1880 does not know about the internet.
This is by design. It's how the past actually worked. People in the past didn't know their own future. If a character could tell you "how it all turned out," they wouldn't be a person of their time, they'd be a modern narrator wearing a costume. Airtrek's characters are people of their time, and that means their knowledge has edges.
When you talk with a character, the conversation is remembered. If you come back and talk to the same character again, they recall what you discussed. If you meet a character who knows one you've already spoken with, they may reference things you said. This is persistent memory, it's what makes the world feel inhabited rather than reset every time you open it.
Memory is stored server-side and is tied to your account. You can read what's stored and delete it. See Historical Accuracy for how memory relates to consistency, and our Privacy Policy for the data details.
Each character has a distinct voice generated by a text-to-speech model. The voices are designed to fit the character, gendered appropriately, pitched to their personality. When a character speaks, the voice is synthesized from their text response.
Some character introductions are pre-rendered and cached so they load instantly and cost nothing to play. Dynamic conversation responses are synthesized live. If the voice service is unavailable, the experience falls back to your browser's built-in speech synthesis so you still hear the character, we don't silently degrade the experience without telling you.
Ask a senator, a noblewoman, and a merchant in Imperial Rome the same question and you get three different answers, because they lived three different lives:
Quintus Aelius Severus
Senator
"The emperor's power is an instrument. Under Trajan it is a foundation. The office is the same. The man is not."

Julia
Noblewoman
"Power is not a tool to me. It is the weather I host my guests in. It decides whether a woman like me may read Greek philosophy without scandal."

Gaius
Merchant
"I'm richer than half the Senate, but I can't hold the highest offices because I'm equestrian, not patrician. The emperor's power is the ceiling I hit every day, no matter how much I've built."
This is not three attempts at the same correct answer. It's three genuinely different answers, because they're three genuinely different people.
Immersive History →
What immersive history is and who it's for.
AI Historical Figures →
The difference between a static chatbot and a multi-perspective experience.
Explore Imperial Rome →
A specific experience page with character examples.
Historical Accuracy →
How Airtrek handles evidence, anachronism, and creative reconstruction.
Join the beta waitlist for early access to Airtrek AI
Join the beta waitlist →