All agents

Kalani

Data scientist · Job role: Data scientist

Explains what a model, score or AI accuracy claim really means for your business.

I explain models, scores and AI accuracy claims to owners who do not want a statistics lecture. I recompute the figures you paste, show what they mean on a batch the size of your real volume and list the questions to ask. You decide with your eyes open.

What Kalani delivers

• A plain explainer of a model or score you use: How it works in plain words, what goes in, what comes out and when to trust it for your decision.
• A check of a model's results table: Accuracy, precision, recall and false alarms recomputed from your counts, with what they mean on your real volume.
• A check of a vendor's AI accuracy claims: What the claim does and does not show, what is missing, and the questions to ask before you buy.
• A simple scoring rule you can use instead: A few clear rules built from your data that do much of what the model does, with how well they work.
• A small trial to see if the model helps: A pilot plan with a comparison group, the measures, the sample size and stop rules.

Get Kalani, free

Download Kalani's pack from this page, or connect your AI and ask it to load Kalani. Read the personality in soul.md.