The starting point. I began building mess.vn in December 2025 for a personal need: a Vietnamese roleplay experience with the familiarity of a messaging app. After sharing it with friends, I released a free MVP to a small roleplay community before the public launch in March 2026.

Learning with the community. I asked early testers short, open-ended questions and offered in-app coins for feedback. Direct messages, Discord discussions and later incentivized polls helped surface requests and friction, guiding ongoing product iteration.

Built with Claude, directed by me. I designed and built mess.vn end to end using Claude-assisted development, owning product direction, UI, implementation and ongoing iteration.

Design decisions.

• A familiar messaging interface. Conversation layout and typing indicators borrow familiar messaging patterns to make the roleplay experience feel natural.
• Memory for continuity. Remembering details from earlier conversations was designed to make each session feel connected, instead of starting over.
• Emotion turned into mechanics. Affection, trust and jealousy read as game stats. Every conversation becomes something you can play, and want to win.
• Designed for revenue. I built the M Coins economy like a small fintech product: top-up flow, balance, transaction history, daily streaks, reward milestones and the Scarlet membership tier.

The outcome. The product reached 100K+ users, approximately 5,000 peak daily active users and around VND 300M in peak monthly revenue, with no paid advertising. I balanced AI operating costs, usage limits and paid access to support profitable operations. These are product-level results, not a measured conversion lift from a single design change.