I turn messy real-world data into clean, structured, ready-to-use datasets. My focus: computer vision + geodata — pointing modern AI models at street-level imagery so you can query the physical world like a database.

Google Street View - Semantic Search  Describe what you're looking for — "red storefront", "outdoor cafe seating", "construction site" — and get every matching Google Street View image within a radius, scored by AI, geotagged, and streamed live to your dataset.

• 🧠 CLIP visual understanding: concepts, not keywords
• 🗺️ Works anywhere Street View has coverage
• ⚡ Zero setup — no Google API key, no billing account
• 📅 Date filtering: only imagery newer than your cutoff

Need a variation — different imagery source, custom scoring model, bulk area scans? Open an issue on any of my actors or reach out through Apify.