LocalWhisper | Transcriber & Summarizer: Turn raw transcripts into private business briefs via Llama3

Hey Product Hunt! 👋 I’m the maker behind LocalWhisper Business Brief Generator. If you use local Whisper models (via Ollama, MacWhisper, or CLI), you know the pain: the audio transcription is great, but the actual text output is a messy wall of "uhs", self-corrections, lost action items, and unstructured numbers. Reading through a 30-minute transcript to find actual decisions takes forever. Sending internal business transcripts, financial discussions, or strategy notes to cloud APIs for cleanup wasn't an option for me — data privacy comes first. So I spent the last few days building a native, 100% offline solution for macOS. What LocalWhisper actually does: Filters Speech Noise: Strips disfluencies, filler words, and conversational artifacts. Extracts Signals from Noise: Isolates explicit decisions vs. casual ideas. Auto-generates Structured Briefs: Outputs clean Markdown with Executive Summaries, Timelines, Risks, and Action Item tables (with owners & deadlines). Detects Contradictions: Flags conflicting statements made during the discussion. 100% Local & Zero Cloud: No network calls, no telemetry, no data leaving your Mac. Tech Stack: Built with Tauri 2 (Rust) for an ultra-lightweight macOS desktop UI, powered by a zero-dependency Python (stdlib only) engine under the hood. Fast, lightweight, and completely isolated. Would love to get your feedback, feature requests, or answer any technical questions about the build! 🚀