Three layers, one descent
You always know which depth you are at, what you can click into, and where our coverage stops. The trail reads Europe → Poland → region → parcel and every step of it is clickable back up.
- Europe
- Poland
- Mazovia
- Parcel
Market
Which country or region should we be in?
Compare European markets on available and forecast grid capacity, typical connection wait, power price and generation mix, renewable share, land cost, permitting timeline and regime, regulatory burden including efficiency and waste-heat obligations, water availability, fibre and latency position, tax and incentive regime, and the political stability of that regime.
Weighting is the product, not a filter
A fund weighting power price against permitting speed reaches a different ranking than a developer with a signed offtake. You set the weights; the ranking and the map recompute together, and the weighting travels with anything you export.
Rank everything, or compare three
Two real questions, two views of the same data: rank all markets on my weighting, or put these three side by side on every criterion with the sources for each.
Area
Where inside that market?
Within a chosen country: a regional capacity heat map, transmission corridors and the substations with headroom, fibre routes, water availability, the zones actively courting data centres, and regional grid expansion plans. Layers toggle independently, and the output is a shortlist of areas you can save — not a verdict.
- CAPACITY
- TRANSMISSION
- SUBSTATIONS
- FIBRE
- WATER
- COURTING ZONES
- EXPANSION PLANS
- Pilot region, site depth
- Area depth only
Poland at launch. Further countries are added when market-depth usage justifies them, so expansion is demand-led rather than guessed.
Site
Will this specific site work?
Enter an address or parcel ID, or draw the boundary — real sites are rarely a single parcel, so multi-parcel selection is standard. You get a fixed field set across 12 categories, each field answered from public data with a confidence state, a named source and a retrieval date.
Site header
Identity, geometry, and a completeness summary: how much of the field set we filled, and how the states distribute.
Twelve categories
Power, zoning, environment, flood, water, fibre, access, land conversion, title, geotechnics, tax and local politics — each with its own state distribution.
Every field, expanded
Source, link, retrieval date and the quoted clause. Red flags and open items stay one click away from anywhere in the document.
Power
Capacity, connection point, voltage level, queue position
Zoning & planning
Instrument in force, permitted use, height, coverage
Environment
Protected areas, habitat, EIA trigger, screening
Flood
Q1% and Q10% extents, intersection area, defences
Water
Abstraction, discharge consent, cooling volume
Fibre & latency
Routes, carriers, IX distance, measured latency
Access & transport
Road class, load limits, rail, construction access
Land conversion
Agricultural or forest class, conversion route, KOWR
Title & rights
Ownership, encumbrances, easements, pre-emption
Geotechnics
Ground conditions, bearing, mining subsidence
Tax & incentives
SEZ status, property tax, grant eligibility
Local politics
Council record, opposition, litigation, moratoria
Geometry, computed not eyeballed
Buildable area after setbacks, slope, flood intersection, and measured distances to transmission, the nearest substation with headroom, fibre, road and water. The buildable polygon is drawn distinct from the parcel boundary, because the difference between them is usually the story.
Reading documents, not just maps
The highest-value findings in testing came from regulatory and news documents rather than GIS. Pass one retrieves the applicable zoning instrument and answers a fixed question set with clause citations. Pass two reads local news, council minutes and permit registers for opposition, litigation, moratoria and plan revisions. Site-killing risk is often political rather than physical, and it is public. Both passes can answer "not determinable from available documents".
A run names each source as it is queried, and a source that fails or times out is recorded rather than skipped. Its fields carry the outage and its date into the dossier.
- PSE transmission networkOK
- MPZP · gmina NadarzynOK · 2026-08-14
- ISOK flood zonesNO RESPONSE · 3 TO CONFIRM
- UKE / SIIS fibreOK
- Wody Polskie · abstractionQUERYING…
Nobody should discover a wall by hitting it
Coverage depth per country is a first-class attribute, visible on the row and on the map before you click. Where we do not yet go deeper, that is a designed answer: what we do have for that market, what it would take to build the next depth, and a way to register interest.
Poland is where the method proves out. The source documents are Polish, the plans are municipal and five separate DSOs answer for the grid, so nothing about it is shallow work. It is also tractable: the data is national rather than regional, there is no cadastre licensing to negotiate, and we can read and verify every source ourselves. Usage in the upper layers decides which country we deepen next.
A first-class deliverable, not an export
Most real use is a document forwarded to someone who never opens the app — an investment committee, a lender, a client. So it stands alone: cover page with site identity and completeness, executive summary with red flags and open items up front, full field tables with provenance and citations, map plates, and a sources appendix with retrieval dates.
Print-safe throughout. The state indicators survive black-and-white printing and photocopying, which is how these documents actually travel.
- Cover · site identity, completeness summary
- Executive summary · red flags and open items
- Field tables · provenance and citations, 12 categories
- Map plates · parcel, buildable area, constraint layers
- Sources appendix · every source, every retrieval date
