WingMan AI — Dating Coach

WingMan AI is a mobile-first conversation coach for dating apps. Users can upload a chat screenshot or enter text, review an analysis view, and draft or refine replies in a chosen tone.

Responding thoughtfully to a message can be difficult when the context is ambiguous or the user is unsure how to continue the conversation. The concept explores a guided workflow that helps users review context and consider possible responses without treating an automated interpretation as certainty.

The interface is organized around four main tasks:

Capture context: upload or paste a chat screenshot, or enter conversation text.

Review signals: inspect an analysis view designed to surface possible positive and negative cues.

Workshop a reply: revise a draft into a preferred tone.

Plan a follow-up: use a strategy-coaching view to develop the next question or response.

React 19, Vite, and Tailwind CSS 4 for the front end.

Express server with request validation and server-side Gemini API calls, keeping the model key out of the browser.

Custom instructions to carry a preferred tone across requests.

Responsive layouts, onboarding, Android-style navigation, a PWA manifest and service worker, and Play Store packaging assets.

The application was run locally; the interface loaded and navigation worked.

Gemini-powered image/text analysis, reply generation, and coaching were not run because no API key was available. Output quality remains unverified.

There is no automated test suite.

Chat screenshots may contain personal information. A real deployment would require privacy review, secure handling, and explicit data-retention rules.