Bready

Bready is a TypeScript / Electron desktop AI interview copilot. It supports both interview preparation and live interview assistance, using voice or text prompts and resume/JD context to produce answers that fit the candidate’s real background.

I built the 0→1 product concept and the full-stack desktop architecture independently. The project uses Electron, React, Express / WebSocket, PostgreSQL, JWT / email authentication, and model integrations such as Doubao and Gemini.

The product supports Chinese, English, mixed Chinese-English, Japanese, and French interview scenarios. Its privacy framing combines local runtime and a relay-style backend: user data is not stored as a platform asset and is only forwarded when a configured model service is called.

The public repository shows the desktop structure, UI, database layer, and diagnostics, but the product has not been formally released or promoted. This page therefore describes functions, architecture, and status without adoption or commercialization metrics.