AI with a visual interface
The Hook
“When I tried ChatGPT for the first time, like everyone else, I was blown away. Being able to ask any question and getting an intelligent answer was nothing short of groundbreaking. But the user interface felt like we had gone all the way back to the command line days.”
Problem Agitation
“I was at Airbnb at the time, and it was clear that chat wasn't the right interface for travel or many other use cases. We needed something more visual and interactive. Fast forward a few years, I was talking to our cofounder Murat, and he had this idea: What if AI could generate the entire user interface based on your question? But for a magical user experience, this had to be done in just a few seconds.”
Product Reveal
“It definitely wasn't easy, but we think we made it work. Let me show you Monogram: the first AI with an intuitive visual interface. I promise it will make sense when you see it.”
Feature Teaser
“Let's start with a simple example. Can you show me all the TV shows which have a new season coming out this year? Monogram is getting data from multiple sources, and when ready, it generates a visual response in just a few seconds... In this case, the details of the TV show... Monogram even decided to add a watch button... It can deep link into content accurately or send you to a website.”
Feature Teaser
“Can you give me a recipe for Turkish stuffed eggplants, but use lentils instead of ground beef? Let's give it a couple seconds. This time, it creates an interface with ingredients, step-by-step instructions, and more things I can do, like creating a shopping list or modifying this recipe.”
Feature Teaser
“Can you help me schedule a trip to Yosemite this weekend with my kids? This time Monogram decided to verify some details... Our interaction model blends natural language with more familiar UI affordances... Option pickers reduce your cognitive load of decisions... Now it's showing me some hotels I can stay, some outing ideas, scenic anchors, and a day-by-day trip plan.”
Call to Action
“It makes a big difference when AI can generate interfaces on the fly. Just like graphical user interfaces did for personal computers, we believe that making AI easier to use will help it benefit more people.”
Synthesizable in Remotion · key components of the keynote-device-split-screen template.
Presenter pinned to left third while interactive 3D phone device viewport floats on right.
Layout component with flex container; left side renders presenter video asset, right side applies interpolate 3D rotation and spring-animated phone mockup.
Dynamic vertical list of cards smoothly staggered in on API response completion.
Map over results array with Spring physics on translateY and opacity staggered by frame index * 3.
Modal picker sliding up from device bottom for calendar dates and chip selections.
TranslateY animated from 100% to 0% with spring dampening 25 and stiffness 120 over background overlay fade.
Purpose. Highlight friction and visual limitations of standard conversational LLM chat.
Execution. Presenter speaking directly to camera on solid indigo studio set.
Purpose. Establish generative on-the-fly UI as the natural evolution of AI interaction.
Execution. CPO interview footage intercut with floating 3D device visual.
Purpose. Demonstrate breadth across media recommendations, complex culinary recipes, and travel itinerary generation.
Execution. Split-screen synchronized voice queries with live responsive mobile screen recordings.
Purpose. Compare generative UI to the historical invention of the GUI.
Execution. Slow camera push-in on founder followed by minimal monogram.ai lockup.