Proprietary development
SoulGCI
Coordinates AI and expert work from task to result.
← Homepage
TerraScore / Technologies
Technology foundation of TerraScore
Our advantage is the technology developed by the TF HOLDING and TF Intelligence Platform group, connecting long-term data, scientific calculations, scenario modelling and governed AI. This foundation supports new TerraScore tools for land, agriculture, climate and institutional analysis.
02 / TerraScore
Technology foundation of TerraScore
4 patent applications · Monte Carlo engines · Long-term data
Proprietary development
SentientROUTER
Routes AI models for analysis tasks.
Proprietary development
ProvableCORE
Links results to their sources and evidence.
Research
TEHNOFARM OS
A shared architecture for sector-specific systems.
Data Integration
Official, satellite and scientific sources, plus authorized local measurements.
Intelligent Processing
Validation, harmonization, scientific calculations, models and scenario analysis.
Scientific Verification
Data provenance, quality control and human expert review.
Applied Intelligence
Territorial passports, maps, risk analysis and professional reports.
Group technologies support TerraScore’s development. Available tools and data are described in each national portal.
Technology assets for new capabilities
Monte Carlo
Monte Carlo engines and scenario analysis
The group has its own Rust simulator with parallel Rayon processing. Its per-scenario engine models combinations of yield, prices, costs and weather stress to explore a range of possible outcomes. The service operates within the group’s technology environment.
The portfolio programme includes the canonical Python v1.3.0 calculation and a developing Rust kernel with reproducible random sampling, correlations, percentiles, DSCR and IRR. These assets support future comparisons of business scenarios, sensitivity analysis and joint analysis of multiple territories or projects.
The separate Rust service exists; Rust portfolio integration is in development. Scenario estimates depend on their inputs and are not a credit rating.
NASA POWER / MERRA-2
Geographic reach and historical depth
The group’s analytical layers span up to 29 European countries, with scope and completeness varying by dataset. Europe brings together 29 national directions, four published national portals and 41 analytical directions. Prepared datasets, an accessible portal and an operational analytical tool each have their own coverage.
Examples of depth include Italy’s 1991–2025 climate history: 420 source months and a 1991–2020 climate reference period. Preserved analytical datasets include climate series from 2000 and satellite NDVI from 2016. Such series support studying change over time, comparing territories and developing scenarios.
The Italy example uses NASA POWER/MERRA-2 model data. Period, resolution and completeness are checked per source; an incomplete current year is not counted as a full year of observations.
TF HOLDING S.R.L. / AGEPI
Proprietary developments and patent applications
AGEPI documents confirm four patent applications filed by TF HOLDING S.R.L. They document the group’s research and engineering assets in governed AI, verifiable execution and operating architecture.
| Development | AGEPI application | Filing date |
|---|---|---|
| SoulGCI | a 2026 0020 | |
| SentientROUTER | a 2026 0011 | |
| ProvableCORE | a 2026 0010 | |
| TEHNOFARM OS | a 2026 0014 |
The status shown here is documented filing. Each application’s scope is defined by its documents; the TFOS application concerns medical autonomy in isolated environments.