1. Identification
We start from the exact person, company, property and operation. We normalize names and match legitimate identifiers to reduce homonyms before attributing information.
2. Proportional search
We define sources according to purpose and risk. We do not collect data out of curiosity nor do we promise access to confidential systems. We record coverage, date and restrictions of each consultation.
3. Verification and cross-validation
We prioritize documents and primary sources. We compare matches with location, dates, procedural nature, folios, RFC or other pertinent data. A secondary mention never replaces the document that supports it.
4. Contextualization
We separate demand for judgment, presumed final publication, writing of registration certificate and nominal coincidence of confirmed identity. We incorporate clarifications from the owner when appropriate.
5. Prudent classification
We do not use automatic reliability scoring. Each point is classified as confirmed, consistent, requires context, unconfirmed, or insufficient information. The relevance is explained in relation to the specific decision.
- Confirmed: source and identity sufficiently corroborated.
- Consistent: documents and sources do not show material contradiction.
- Requires context: a real event whose meaning depends on additional information.
- Unconfirmed: match or reference without sufficient elements.
- Insufficient information: coverage or support does not allow conclusion.
6. Report and human review
The report presents facts, sources, cutoff date, limits and possible mitigations. It does not accept or reject people; The decision remains with the client and must avoid discriminatory criteria or criteria unrelated to the operation.
Intelligence Center