Hermes Wiki

MachineLearning

Classical ML/MLOps — distinct from AI/ which is LLM/agentic-focused.

Why we need this / what value this brings

Distinct skillset and infra from LLM/agentic work (AI/) — relevant only once there's an actual predictive/classification use case.

When to use this

Once there's a concrete use case (ranking, fraud detection, recommendation) that a rules-based approach can't handle well.

How to use or implement this

Start with the simplest model that could plausibly work (or a heuristic) before reaching for anything sophisticated.

Subtopics

Research questions

  • Does Localz have any actual ML use case yet (recommendation ranking, fraud detection) or is this purely background study?

Empty folder — drop notes, links, and findings here as you research.

Hermes Wiki