---
title: Programme World Model
emoji: 🧭
colorFrom: green
colorTo: blue
sdk: static
app_file: index.html
pinned: true
short_description: Evidence-grounded AI for complex programmes
---

# Programme World Model

A research and engineering project by **Bruno Chikeka**, exploring
evidence-grounded AI for complex programmes.

This Space presents the project’s approach, current implementation and research
limits. It is a static overview, not a trained forecasting model or an interactive
simulation service.

The current implementation includes a public-report evidence pipeline, a
versioned SQLite index, resumable acquisition, source-review and reconciliation
records, backup/restore, and a separate synthetic simulation kernel.

The initial pilot covers 15 implementation reports for World Bank project
P126507. It retains original document hashes and cited observations. Source
coverage, historical availability and independent extraction accuracy remain
under investigation. No model is empirically calibrated.

Only this overview is published here. The implementation repository and source
corpus are not included in this Space.

Profile: https://huggingface.co/Chikeka