I want to add smart features, but I don't want to learn Python just for this.
Setting up a Python microservice feels like overkill for one feature.
I know Ruby can do ML, but I don't know where to start.
What Python devs don't tell you
┌─────────────────────────────────┐ │ Your Rails App │ ├─────────────────────────────────┤ │ Ruby Gem (torch-rb, etc.) │ ├─────────────────────────────────┤ │ FFI/Rice Binding Layer │ ├─────────────────────────────────┤ │ C++ Library (LibTorch, etc.) │ │ ← Same code Python uses! → │ └─────────────────────────────────┘
Ruby ML gems wrap the exact same C++ libraries as Python. When you call Torch.tensor([1,2,3]) in Ruby, you're running the same compiled code.
Your existing stack. No Python required.
Find content by meaning, not keywords.
"Users who liked this also liked..."
Categorize content, detect spam, route tickets.
Predict demand, detect anomalies.
RAG pipelines and embeddings in Rails.
Background jobs, caching, monitoring.
12 Modules. 40+ Hours.
One person has single-handedly made Ruby ML possible.
Be first in line when the curtain drops.