5 ms·

Side project: A native macOS app in Swift that runs locally and uses AI to clean and organize files by moving them into the best-matching folders. No backend or

Side project: A native macOS app in Swift that runs locally and uses AI to clean and organize files by moving them into the best-matching folders. No backend or accounts. https://floxtop.com https://floxtop.com Full-time: C++ work on nearby connectivity (bluetooth) for embedded / industrial devices (factory equipment). Deep stack, hardware constraints, long lifecycles, high reliability. Non-web work feels very different: stronger constraints, slower but deliberate releases, and bugs are much more expensive. There’s a lot of interesting software being built far away from HTTP and browsers.

cpursley 9mo ago
That's neat, using Apple Foundation Models or something else? I'm very curious about how it's determining folder matches (I need to do something for images that are already classified/tagged via FastVLM) in iOS.
- bobnarizes 9mo ago
  Not Apple Foundation Models — unfortunately they’re not capable enough (yet) for understanding content and matching it to folders. I’m using SBERT-style embedding models for the semantic matching, which works very well in practice. For non-text content, the app also analyzes images (OCR + object recognition) using Apple’s Vision framework. That part is surprisingly powerful, especially on Apple Silicon. > I need to do something for images that are already classified/tagged via FastVLM What’s the concrete use case you’re targeting with this?
  - cpursley 9mo ago
    Classifying real estate / property images. Also using Apple Vision which ain't half-bad for something on device and feeding that metadata along with what FastVLM returns into Foundation model to turn into structured output - trying to see how far a I can push that. But feels pretty limited/dated in term of capabilities vs lead edge models.
    - bobnarizes 9mo ago
      I’ve seen a huge advantage in running everything fully local and private. Not sure if that fits your use case, though. Nearly 90% of Floxtop users choose the app mainly for that privacy focus.

cpursley 9mo ago

That's neat, using Apple Foundation Models or something else? I'm very curious about how it's determining folder matches (I need to do something for images that are already classified/tagged via FastVLM) in iOS.

bobnarizes 9mo ago
Not Apple Foundation Models — unfortunately they’re not capable enough (yet) for understanding content and matching it to folders. I’m using SBERT-style embedding models for the semantic matching, which works very well in practice. For non-text content, the app also analyzes images (OCR + object recognition) using Apple’s Vision framework. That part is surprisingly powerful, especially on Apple Silicon. > I need to do something for images that are already classified/tagged via FastVLM What’s the concrete use case you’re targeting with this?
- cpursley 9mo ago
  Classifying real estate / property images. Also using Apple Vision which ain't half-bad for something on device and feeding that metadata along with what FastVLM returns into Foundation model to turn into structured output - trying to see how far a I can push that. But feels pretty limited/dated in term of capabilities vs lead edge models.
  - bobnarizes 9mo ago
    I’ve seen a huge advantage in running everything fully local and private. Not sure if that fits your use case, though. Nearly 90% of Floxtop users choose the app mainly for that privacy focus.