Wingman is an AI-based dating application built as a direct reaction to the dominant mechanics of modern dating apps.
Most dating platforms optimize for engagement, not outcomes. Infinite swiping, shallow signals, and gamified feedback loops produce activity, but not necessarily connection. Wingman starts from the opposite assumption: less choice, more context, better intent.
The app removes swiping entirely. Instead, users go through a structured onboarding that captures personality traits, values, preferences, communication style, and expectations. This information is stored as a rich, flexible profile rather than a fixed schema, allowing the system to reason about compatibility instead of simple filters.
Matching in Wingman is curated and limited. Users don’t browse hundreds of profiles. They receive a small number of high-quality matches computed by the system, with clear explanations of why a match exists. The goal is to shift the mental model from “shopping” to “considering”.
A defining feature is the concept of introductions. Conversations don’t start by default. When two users are matched, the AI can generate a contextual introduction — a short, meaningful opener grounded in shared traits or values. This lowers the activation energy while avoiding generic icebreakers.
From a technical standpoint, Wingman is designed around flexibility and iteration:
- Firebase / Firestore stores highly structured but evolving user profiles in JSON form
- Cloud Functions handle matchmaking, daily match computation, and introduction logic
- OpenAI models are used for personality synthesis, compatibility reasoning, and message generation
- The mobile app acts primarily as a client to a rules-driven backend rather than embedding logic locally
One important constraint is intentional friction. Wingman is not designed to maximize daily usage. Limits on matches, introductions, and interactions are deliberate. Scarcity is used to increase attention, not addiction.
On a personal level, Wingman is the most opinionated project I’ve built. It encodes a clear stance: technology should reduce noise, not amplify it. Building it forced me to formalize assumptions about relationships, communication, and how incentives shape behavior — and then translate those assumptions into product constraints.
Wingman is still in development, but its direction is stable. It’s not trying to win the dating app market. It’s trying to build a different category altogether: slower, more deliberate, and aligned with how meaningful relationships actually start.
