Chungin Roy LeeCEO & Co-founder

We built a small demo called Interview Coder. It was a clear pane over your apps that could see your screen and feed you answers. The demo went viral. I used it in an Amazon interview, posted it, VCs saw it, and we raised fast. Then we went all-in on Cluely.

It sits on top of your computer like a glass sheet. It sees your screen, hears your meeting, knows context about you, and predicts what you need. Instead of you asking a chatbot, it surfaces the right info in real time.

People already use AI for essays, homework, even tests. If the world says that is “cheating,” then I might as well embrace it. Saying “we are not a cheating app” while users do it anyway is worse. Owning it gives us a marketing edge.

Interviews and enterprise. Call centers used to memorize 600-page manuals. Now you dump that into AI and get live guidance. Sales reps can answer technical questions on the fly. It feels like you studied for hours.

Models are strong, but the magic is context stitching. We convert what you see and hear into optimal text tokens the model can reason over. For many tasks we use the latest state-of-the-art model, but our input pipeline is the edge.

Best today is probably Grok 4. We are not “better than Grok.” We are better than the user at prompting and packaging context.

The goal is “AI always with you,” but Apple will not let an app watch your Snapchat or Instagram like that. Desktop lets us see meetings, the browser, the workflow.

Short-form is the biggest opportunity of our lifetime. You can speak for 10 seconds and reach 100 million people. If a product is inherently viral, it can carry a company. I doubled down when I saw the first post take off.

I used college as social lubricant. I went to find a co-founder and a wife. I talked to 50 smart kids. One said yes. As soon as I had what I came for, I left to build.

We are very generous with equity. Early team got single-digit percent each. We set a 20% option pool early. Founders still hold majority. I want people to feel like owners.

Move fast in the next six months. Half engineering, half marketing. We will do stunts that create cultural moments, like a huge rave and even a full anime, because attention compounds.

People want not to be lonely. If the team is in the office 12 hours a day, I cannot send them to cold approach. Hinge Premium is a simple way to support their real life. Happier people build better product.

Build something with AI and push it with short-form. If you cannot code, go to hackathons and find an engineer. Or use AI builders to ship a simple app fast, then test content until something hits.

Winner’s effect. One win makes the next win easier. Take action until you get the first clear win, then double down. I value ego in the sense of declaring a big goal and working hard for it.

Most meetings for support and sales get automated. Research gets pulled up live. If nothing useful is needed from you, you should not be in the meeting. We want to unite human memory in real time.

Compute and data. Bigger context and more GPUs change the game. Privacy rules matter, but the geopolitical race is real. I am optimistic about ASI. The only scary case is misaligned objectives.

Low social anxiety, willingness to experiment, and hard work. I try things most people will not. I ignore noise, trust my own eyes, and keep shipping until something clearly works.