Qingyuan Yang's profile on Product Hunt

About

Helping companies hire smarter with AI. Currently building Aural, an AI-powered interview platform that automates candidate screening through voice and video interviews. Previously: • Associate Partner @ McKinsey — Led AI transformation projects across battery, electronics, and retail industries • Data Science @ Uber — Built ML systems for efficiency improvement Education: MBA @ Georgetown, BS Applied Mathematics @ Wuhan University Technical: Software Engineering, AI/ML, Cloud Based in Hong Kong 🇭🇰 🚀 Currently in beta: aural-ai.com

Badges

Tastemaker

Gone streaking

Gone streaking 5

Maker History

AuralAI conducts the interviews. You get the insights.
reddit-skillsAI agent skills for automating Reddit via your real browser
Apr 2026
AuralOpen-source AI that conducts interviews for you
Apr 2026

🎉

Joined Product HuntFebruary 25th, 2026

Forums

•

3d ago

Aural - Open-source AI interviews over voice, chat, and video

Aural is now open source under the MIT license. Design adaptive interviews, share a link, and let Aural conduct the conversation over voice, chat, or video, ask contextual follow-ups, and generate structured reports. Run the managed service or self-host the full Next.js, Supabase/PostgreSQL platform with Docker and pluggable LLM providers.

•

3mo ago

Aural - Open-source AI that conducts interviews for you

Aural is an open-source AI interview platform. Describe what you want to learn, share a link, and AI conducts the interview over voice, chat, or video — asking questions, following up intelligently, and generating a detailed report with scores when it's done. Built-in code editor, whiteboard, anti-cheating, resume parsing, and pluggable LLMs (OpenAI, Kimi, MiniMax). MIT-licensed and fully self-hostable.

•

3mo ago

reddit-skills - AI agent skills for automating Reddit via your real browser

Open-source Python toolkit that lets AI agents (Claude Code, Cursor, OpenClaw) automate Reddit through your real browser. A Chrome extension bridge communicates via WebSockets with a local Python server. No API keys or OAuth — just your normal browser session. 5 skill domains: auth, publishing, search, social interaction, and compound ops. Full CLI with JSON output. Handles Reddit's Shadow DOM web components. Python 3.11+, MIT licensed.