​
Introduction

Drop in a dataset → get production-grade data science back.
AIUS is an autonomous AI data scientist. Point it at your dataset and your brief, and it runs the full analysis pipeline — profiling, cleaning, modeling, reporting — pausing at review gates so you stay in control of every decision.

​
How you use AIUS

  • The aius CLI — a terminal app where your analyses run. It works on a local project directory (your brief + your data) and walks the pipeline end to end.
  • The Dashboard at aius.co/account — your account, API keys, billing, projects, and the deliverables your runs produce.
  • These docs — guides and the API reference.

​
What AIUS does

AIUS runs the full data science loop autonomously:
  • You provide data and a brief → drop your dataset and goals into a project directory
  • AIUS maps & executes → profiles the data, cleans it, generates and runs code, analyzes
  • You review at the gates → approve its understanding of your brief, the data findings, and the goals before the heavy work starts
  • AIUS packages outputs → reproducible Jupyter notebooks, processed datasets, models, and visualizations
  • AIUS delivers insight → reports, model cards, decks, and briefs published to your dashboard

​
Why AIUS

  • Reproducible by default — every analysis is a real notebook with a full trace. Re-run anything, any time.
  • You stay in control — hard review gates before goals are pursued; permission prompts before anything touches your machine.
  • Auditable — every step of how an insight was produced is recorded, from raw data to final output.
  • Brief-aware — AIUS works from your context and goals, not generic agent defaults.
  • Deliverables, not just notebooks — findings, charts, and summaries packaged for stakeholders.

​
Getting started