Show HN: Audioscrape – From $7 Rust MVP to Podcast Intelligence Platform
Seven months ago, I shared Audioscrape - a podcast exploration tool built entirely in Rust, running on a $7/month VM. Since then, we've transformed it into a full-fledged podcast intelligence platform, serving over 1,000 users, including PR teams, researchers, and marketers.
What's New:
Real-Time Monitoring: Track mentions across the top 100 U.S. podcasts, covering over 80% of U.S. listenership.
Advanced Search: Filter by speaker, sentiment, timeframe, and topic using AI-powered search.
Custom Alerts: Receive notifications for brand, competitor, or topic mentions.
API Access: Integrate podcast monitoring data into your workflows.
Transcription Accuracy: Achieved 92.2% accuracy across 20,000+ episodes.
Technical Stack:
Backend: Axum (async web framework)
Database: SQLite with SQLx for type-safe queries
Authentication: OAuth2
HTML Templating: Askama
Async Runtime: Tokio
Our commitment to Rust has enabled us to maintain low operational costs while scaling effectively.
Try It Out: https://www.audioscrape.com
Discussion Points:
Has anyone else scaled a Rust-based MVP into a production platform?
What strategies have you employed for efficient scaling and user acquisition?
Looking forward to your insights and feedback! Are you using a local Whisper? If yes, what do you use for inference, candle/ort? Not local. Inference is the only part not written in Rust so far. I am using Replicate to run docker images with a pipeline based on faster-whipser, VAD, pyannote and a custom LLM enhancement flow. Thanks for sharing candle/ort. Interesting to see the WASM in-browser opportunities