Public electoral evidence is abundant, but it is distributed across portals, files, manuals, geographic units, election cycles, and changing methodological conventions. Useful territorial intelligence must preserve the connection between a claim, its indicator, transformation, source, territorial vintage, and known limitation.
Hadox Territorial Intelligence is an independent analytical product built around that principle. Mexico is the current reference implementation, not the limit of the product model. The platform is not an official service of, or endorsed by, any electoral authority or public institution.
The working private prototype brings several research surfaces into one governed workflow:
These capabilities support research and decision preparation. They are not causal forecasts, individual-level predictions, or substitutes for qualified human judgment.
The product pattern separates immutable source snapshots, conformed territorial data, analytical features, reviewable model artifacts, and a curated publication package. Public-facing application views receive only the artifacts needed for each surface rather than unrestricted access to the analytical lakehouse.
The next architectural step is to place protected data and calculations behind versioned server-side services. That boundary will support consistent authorization, audit logging, quotas, model-version reporting, and reuse across other Hadox products without exposing internal storage layouts.
The baseline operates on aggregate public evidence. It must not be used to infer individual political behavior, suppress participation, construct sensitive personal profiles, or automate consequential decisions without qualified review. Predictive targets are territorial proxies; scenario benefits and costs are assumptions that require sensitivity analysis.
Production use also requires server-side authorization, hardened identity and recovery flows, least-privilege data access, privacy-aware audit logs, dependency and secret scanning, tested backups, incident response, and independent methodological validation.
The public GitHub repository is a documentation-only disclosure package. It contains the product rationale, capabilities, conceptual architecture, methodology, maturity assessment, limitations, roadmap, governance, security policy, and third-party notices.
It deliberately excludes application source code, commercial interface assets, credentials, user records, infrastructure configuration, private data, source archives, and the full analytical lakehouse. The private prototype remains the working reference implementation.
The project is a working analytical reference prototype, not yet a production-grade multi-tenant intelligence service. Priority work includes a legally clean first-party product boundary, protected service APIs, data contracts and dataset cards, end-to-end authorization testing, observability, independent analytical validation, and a second country implementation.
The public concept material is proprietary and source-visible, not open source. Public visibility does not grant a right to reuse the material beyond the terms in the repository's license. Data publishers, institutions, and third parties retain their respective rights, and references to them do not imply endorsement.