Imouto

Imouto: Multi-Country AI Job Exposure Pipeline

Built the core Imouto pipeline: source-agnostic Python scorer with O*NET and OSCA/JSA data sources, Claude-powered task scoring, US-AU occupation matching, and a SvelteKit frontend.

5 Phases
22 Tasks
2 Days

From Concept to Explorer

Imouto scores every task within an occupation for AI automation exposure using Claude. The pipeline architecture is source-agnostic — pluggable data sources (O*NET for US, OSCA/JSA for Australia) feed a common scorer that produces a dataset with per-task scores, AI tool suggestions, and rationale.

Multi-Country Support

The dataset schema (v2.0) supports multiple countries with country, source, native_code, and matched_id fields. The US-AU occupation matcher uses Claude for semantic matching, enabling cross-country comparison views.

SvelteKit Frontend

An interactive explorer lets users search occupations, view task breakdowns with colour-coded automation levels, and compare US/AU equivalents side by side.

Features Delivered

Pipeline Architecture

  • Source-agnostic scorer — Pluggable data sources with common scoring interface

Australian Data

  • OSCA and JSA data sources — AU occupation data with JSA pre-score bypass

Frontend

  • Job Explorer and Task Breakdown — SvelteKit app with search, country toggle, and task details