Randy Olson

@randal_olson

AI researcher & builder. 15+ years building AI. Data viz aficionado. Building things independently for now.

following
followers

How to use decision models for AI evals

Hamel Husain and Antaripa Saha compare models on 100 leasing-agent traces, warning against an unchecked 0.50 threshold.

12 Sources, 3h ago, updated 3h ago

evident-charts is an open-source agent skill for data visualization

Its creator says the skill works in Claude Code, Codex, Cursor and other tools, drawing on research and expert consensus about effective data visualization.

2 Sources, 24d ago, first seen 24d ago

Auditing AI evaluation pipelines with evals-skills

A co-creator recommends running /eval-audit on an existing pipeline, saying people often find easy opportunities for improvement.

2 Sources, 25d ago, updated 25d ago

Featured in

How to use decision models for AI evals

evident-charts is an open-source agent skill for data visualization

Auditing AI evaluation pipelines with evals-skills

Featured in

How to use decision models for AI evals

evident-charts is an open-source agent skill for data visualization

Auditing AI evaluation pipelines with evals-skills