Structured, discoverable public data is the raw material for artificial intelligence applications, and that is the explicit intent here. The portal is linked to data-heavy uses in agriculture, healthcare, climate, legal research and financial inclusion — domains where models are only as good as the datasets available to train and validate them. Fragmentation across 458 repositories is precisely the condition that makes such work expensive: most of the effort goes into finding and reconciling data rather than analysing it. Making the aggregation open and non-commercial also matters for who benefits, since a paywalled version of the same index would advantage the organisations that least need the help.