First data connectors are live
G-NAF, cadastral parcels, Geoscience Australia bushfire boundaries, SILO climate and DEA Coastlines are loaded as parquet on R2 and queryable by the agent.
Private preview · Land spine live
Datasets
The agent for Australia’s open data
Australian governments and regulators publish unusually good data — addresses, parcels, hazards, emissions, registers — and then scatter it across dozens of portals with nothing to join it on. Zenancy harmonises those portals into one lake and puts an agent with a sandbox over it. Every number traces back to a public dataset and a documented method.
Resolve these 240 addresses to parcels, then tell me which sit in a flood planning area, on bushfire-prone land, or within 100 m of a receding coastline.
| Step | Result | Source |
|---|---|---|
| Addresses resolved to a G-NAF point | 236 / 240 | g_naf |
| Points matched to a cadastral parcel | 231 / 236 | cadastre |
| Parcels inside a flood planning area | 31 | nsw_flood_studies |
| Parcels on bushfire-prone land | 18 | ga_bushfire |
| Parcels within 100 m of a receding coastline | 6 | dea_coastlines |
| No published flood study — reported unassessed | 27 | — |
g_nafcadastrensw_flood_studiesga_bushfiredea_coastlinesBuilt on Australia’s open data
G-NAFCadastral parcelsGeoscience Australia bushfire boundariesNSW flood studiesSILO climateCMIP6 (SSP2-4.5 / SSP5-8.5)DEA CoastlinesNGER emissionsACCU register
Browse the data catalogueHow it works
Harmonise the portals, then ask across them
Four steps, all of them inspectable. The agent writes the query, runs it in a sandbox against the lake, and shows its working.
Ingest and harmonise
Pipelines pull each open dataset from its portal, normalise the schema, record the vintage and licence, and land it as parquet in one lake.
Resolve onto a spine
Everything hangs off shared keys rather than fuzzy strings — parcel, G-NAF address, coordinate, H3 cell — which is what makes the joins hold.
Ask in plain English
The agent plans the analysis, writes the SQL and Python, runs it in a sandbox with the lake mounted, checks the result and corrects itself.
Get the working, not just the answer
Output names every dataset, vintage and assumption, and reports unassessed records as unassessed rather than substituting an average.
Spines
Two spines, many verticals on top
A spine is a set of keys enough datasets already carry that everything else can be hung off it. Get the spines right and you can answer joins across portals that no single incumbent can answer.
Live
Land
Cadastral parcel, G-NAF address, coordinate and H3 cell. Hazard layers, climate projections, land use, soil and water sit on top of it, joined at the parcel rather than at a rooftop pin.
Datasets on the land spineComing
Entity
ABN, ACN and directorship records, with the registers, licences, contracts and filings that reference them. It is the second spine because most of the remaining interesting joins are between an entity and a place.
Use cases
One platform, a vertical at a time
Each vertical is the same lake and the same agent with a different expertise pack on top. Climate disclosure is the first one shipped because the deadline is legislated and the data is already public.
First case study · shipped
AASB S2 climate disclosure
Mandatory climate reporting gave several thousand Australian entities a physical risk question they cannot answer from any one portal: flood studies sit with councils and state agencies, bushfire-prone land with another, climate projections with a third, and none of them share a key with a lease schedule. On the land spine it is a single run.
Group 2 entities report from FY2026-27. The output is an AASB S2-shaped table with a methodology appendix naming every dataset, vintage and assumption — the part an auditor asks about.
Score our 240 sites against flood, bushfire, coastal retreat and SSP2/SSP5 heat-days, then output the AASB S2 disclosure table with a methodology appendix.
| Hazard | Sites at risk | 2030 SSP2-4.5 | 2050 SSP5-8.5 | Exposure |
|---|---|---|---|---|
| Riverine flood | 31 / 240 | Moderate | High | $84.2m |
| Bushfire | 18 / 240 | Moderate | High | $41.7m |
| Coastal inundation | 6 / 240 | Low | Moderate | $22.9m |
| Extreme heat (>35°C days) | 112 / 240 | +9 days | +27 days | Operational |
cadastrensw_flood_studiesga_bushfiredea_coastlinescmip6In build
Agriculture, water and carbon
Soil carbon feasibility, ACCU method screening, water entitlement checks and pre-purchase due diligence on rural land — land use, soil, rainfall and registers joined at the parcel.
Agriculture & carbonComing
Procurement and the entity graph
Government contract notices joined to ABN, ACN and directorship records, so a supplier, its related entities and its award history read as one thing.
Coming
Markets and infrastructure
Energy market data against network and generator registers, joined back to the land spine at the parcel.
Coming
Renewables siting
Renewable energy zones, transmission, land use and constraint layers screened together before a site visit is booked.
News
Latest from Zenancy
Australian flood data coverage is the hard part
Flood exposure is the disclosure item reporters worry about most, and the one where Australian open data is least complete. Here is how we handle the gaps.
Why we're building Zenancy
Australian climate risk assessment is priced for large financial institutions. Mid-cap reporters face the same standard on a fraction of the budget, and the data they need is already public.
Request access
Bring us the question no portal answers
We’re onboarding a small group of design partners, starting with Group 2 reporters and the accounting firms advising them. Tell us the question and the data you have, and we’ll run the first one with you.
Request access