← Back to Studio
NVSBL
Autonomous product discovery, API, and subscription platform
A three-repo suite built around a central question: can autonomous agents continuously discover product opportunities and surface the best ones for review? An invisible engine scans and scores signal, a public API packages approved findings, and an authenticated app closes the loop from discovery to delivery.
- HN Scout runs on a schedule — scans Hacker News for complaints and unmet needs, scores 280+ posts per run, and labels validated opportunities as GitHub Issues before human review.
- Discovery is structured through the Double Diamond: agents diverge across signal sources, scoring surfaces candidates, and human review converges on what gets approved and packaged.
- Three separate surfaces with different evidence boundaries: private agent engine, public site + API, and an authenticated product dashboard — each layer exists to close the discovery loop, not just to ship features.