Hey Product Hunt community! 👋 I'm Vishwjeet, creator of WishPilot. Preparing for high-stakes technical drills, system design rounds, and behavioral interviews can be nerve-racking. When I looked at existing AI interview tools, two things shocked me: 1. They charge $50 to $100 every single month just to practice your own skills. 2. They route your live microphone audio to closed cloud servers with 5+ seconds of lag. I built WishPilot to solve this — completely free, local-first, and open source under GNU GPL-3.0. ⚡ What makes WishPilot different? - 100% Free Forever: Zero paywalls, zero subscriptions, no credit card required. - Sub-Second Speech-to-Text: Powered by an AudioWorklet pipeline and Groq Whisper Large v3 Turbo (~180ms - 350ms acoustic transcription). - 9 Unified AI Brains: Real-time streaming via Groq, Cerebras, NVIDIA NIM, Google Gemini, OpenAI, and Together AI. - Screen-Share Safe: Native Windows Display Affinity (WDA_EXCLUDEFROMCAPTURE) keeps the overlay completely invisible to Zoom, Teams, and Google Meet. - 100% Local-First Privacy (BYOK): All API keys, transcripts, and audio buffers remain strictly on your local PC. WishPilot v1.0.0 is available today for Windows. Explore the project: 🌐 Website: https://wishpilot.vercel.app/ 🐙 GitHub: https://github.com/vishwjeet27/w... I'd love your feedback, questions about the architecture, or feature requests. What local models or integrations would you like to see next? Thank you for your support! 🚀

WishPilot

Free, local-first speech AI & interview copilot

Free, local-first speech AI & interview copilot