About Zenancy

Zenancy is an agent over Australia's open data. Ask it a question about a place, an asset or a portfolio, and it works out which public datasets answer it, joins them, and shows you how.

Why we exist

Australian governments and regulators publish unusually good data. Addresses, cadastral parcels, hazard mapping, climate projections, land use, soil, water entitlements, emissions, company and carbon registers — most of it free, most of it well maintained.

It is also almost entirely unjoined. Each agency publishes into its own portal, on its own schedule, in its own schema, with no key that reaches across to any other. A question that is trivially answerable in principle — which of these sites is on bushfire-prone land and which of them are held by the same entity — turns into months of assembly work in practice. So it mostly does not get asked.

That assembly is the product. We ingest each portal, harmonise it, and hang it off a small number of spines so it can be joined: the land spine first — parcel, address, coordinate, H3 cell — and the entity spine, ABN, ACN and directorship, next. On top of those, the joins across portals that no single incumbent can answer become one query.

Verticals, not one product

The first vertical we shipped is AASB S2 climate disclosure, because mandatory climate reporting handed several thousand entities a physical risk question whose data is already public and completely fragmented. Agriculture, water and carbon are close behind. Procurement and the entity graph, market and infrastructure data, and renewables siting follow the entity spine.

Each of those is the same lake and the same agent with a different expertise pack. That is the point of building the spines first.

How we think about it

Traceability over convenience. Every number cites its dataset, vintage and licence. A result you cannot defend in front of an audit committee is not a result.

Gaps are findings. Australian coverage is uneven — hazard mapping especially. We report unassessed records as unassessed instead of quietly substituting an average.

Open data, openly cited. We do not build proprietary models. The value is in the assembly, the method and the interface, not in withholding the inputs.

The stack

An agent with a sandboxed Python and DuckDB environment, a parquet data lake on Cloudflare R2, parcel-level maps from vector tiles, and pipelines written in Python that ingest and harmonise each open dataset. Rows never enter the model's context — results are handled by reference, so summaries are computed rather than guessed.

Where we are

Private preview, with the land spine live and design partners working on the first vertical. If you have a question that spans several Australian portals and nothing to join them on, request access.