Meet Kent
I’m Kent Brought, an undergraduate at MIT studying computer science and comparative media studies and the founder of KayBer and YapClap. I’ve spent the past few years building products that let people meet, talk, and learn together in real time. My background ranges from computer-vision research and educational AI tools to launching full-stack social apps with WebRTC video chat, React, React Native, and moderation models. Along the way I’ve led teams, shipped code to thousands of users, and earned support from programs like MIT Sandbox.
What inspired you to create KayBer and YapClap?
I started YapClap right after Omegle shut down. Watching how many people cared proved there was still huge demand for spontaneous one-on-one chats. With my experience building MacLea, I already knew we could handle the moderation problem with AI instead of giving up on safety. Omegle always felt like a gimmick you tried once for a laugh, but meeting new people is a basic human need, so it should be instant, safe, and actually useful.
KayBer grew out of my love for old school forums like 4chan, where threads felt raw and alive, but they never evolved for mobile, modern UX, or real moderation. I wanted that same energy in a cleaner, faster format. YapClap handles the live face-to-face spark, and a lot of those moments flow back into KayBer when people want to keep the conversation going, post, or share something after they connect.
How does YapClap differ from platforms like Omegle and Chatroulette?
Legacy random-chat sites never solved safety, so users either risked exposure to explicit content or had to self-police. YapClap pairs users in under half a second but screens every video frame before it is sent, so explicit material never even touches the network. We also offer audio-only and text-only modes, interest filters instead of pure randomness, and lightweight reputation scores that steer respectful users toward each other. There is no login requirement, yet you can preserve an anonymous handle if you want to reconnect later.
Who is your ideal user, and how are you reaching them?
The core audience is curious Gen Z and college students who crave authentic conversation but are tired of curated feeds. They’re comfortable with ephemeral content and value privacy. We reach them through short-form video demos on TikTok and YouTube Shorts and campus flyers at colleges where we have early adopters. Word of mouth has been surprisingly strong because a single positive call often turns a first-timer into an evangelist.
How exactly is AI used in KayBer’s matching algorithm?
We calculate a lightweight reputation score based on how often other users like or dislike interactions, adjust for session length, and decay the score over time so new behavior matters more than old mistakes. We combine that score with declared interests, language preference, and device type, then feed the vector into a nearest-neighbor search. The same pipeline powers both video and text matching.
How do you protect users from harassment or inappropriate content?
First, every video frame and text snippet is run through models that detect nudity before anything leaves the user’s phone or browser. If content is flagged, it is blocked locally and the user receives a warning that escalates on repeat offenses. Second, we give both participants an instant block button that immediately ends the session and records a private report. Finally, our reputation score quietly deprioritizes users who collect blocks, so bad actors end up talking mostly to themselves.
What are your thoughts on the growing trend of AI companions?
Digital companions are a useful tool in the same way a journal or a smart speaker is useful. They help with reflection and practice conversations. Long term, though, most people still want the unpredictability and energy that only another human can provide especially in times of true loneliness. That’s why I see AI companions coexisting with KayBer and YapClap rather than replacing them. Our platforms focus on facilitating genuine human encounters while using AI in the background to keep those encounters positive.
What’s the biggest misconception people have about your platform?
Some think anonymity automatically means chaos. In reality, anonymity can encourage honesty and vulnerability when users trust they won’t be doxxed or tracked. Because our moderation runs at the edge and our matching algorithm rewards good behavior, the majority of calls on YapClap and posts on KayBer are friendly and often surprisingly deep. The platform isn’t a wild west; it’s more like an unscripted talk show where guests still abide by studio rules.
Find a painful bottleneck and solve that first before layering on features. For us it was connection speed and safety. Ship a minimal slice, watch real usage, and iterate weekly. Resist vanity metrics like total sign-ups and obsess over retention and average session length instead. Finally, invest early in community guidelines and moderation tech; retrofitting them later is ten times harder.
Did you enjoy our interview? Do you have anything to say to our community?
I enjoyed it. Talking through these questions reminds me why I started building in the first place. To everyone reading, stay curious and keep experimenting with new ways to bring people together online. If you want to see KayBer.com’s social board or try a video chat on YapClap.com, swing by and let me know what you think.
