Clarvia builds source-backed workflow data for bereavement administration.
The methodology is designed to make administrative guidance:
- traceable,
- reviewable,
- reusable,
- multilingual,
- and suitable for human-readable and machine-readable outputs.
Clarvia provides administrative guidance based on official sources.
It does not provide individualized legal advice.
Clarvia's workflow production loop is:
find source
→ create source object
→ extract structured facts
→ attach provenance
→ validate schema
→ human review
→ publish
→ monitor for changes
Every publishable workflow item should be traceable to an official or authoritative source.
Preferred sources include:
- government portals,
- official forms,
- statutes,
- regulations,
- administrative guidance,
- official institutional pages,
- and official service descriptions.
Unsourced administrative, legal, tax, or inheritance claims should not be published.
Clarvia uses a small set of explicit objects:
- Source
- Institution
- DocumentRequirement
- Deadline
- Condition
- Task
- Workflow
- Scenario
- ReviewEvent
This model is intentionally conservative.
The goal is to support useful public outputs without building a heavy case-management or legal-advice system.
Clarvia uses the following verification states:
| State | Meaning |
|---|---|
discovered |
A source or claim has been identified but not yet structured. |
structured-from-source |
A fact has been extracted into structured data from a source. |
source-checked |
A human has checked the structured data against the source. |
expert-reviewed |
A qualified reviewer has reviewed the item. |
published |
A maintainer has approved the item for public output. |
stale-review |
The review is outdated or the source may have changed. |
superseded |
The item has been replaced or is no longer current. |
Only maintainers should mark content as published.
Workflow objects should include:
- source references,
- jurisdiction,
- language,
- access date,
- verification status,
- and review metadata where applicable.
Public outputs should display:
- source citations,
- last-verified dates,
- correction pathways,
- and review or publication status where appropriate.
Human review is required before workflow content is published.
Human review is especially important for:
- deadlines,
- administrative obligations,
- required documents,
- jurisdiction-specific branching,
- cross-border scenarios,
- and inheritance or succession-related workflow logic.
AI systems may assist with:
- source discovery,
- metadata cleanup,
- first-pass extraction,
- translation drafts,
- issue drafting,
- documentation drafts,
- validation support,
- and changelog drafting.
AI output must be treated as draft material.
AI-generated factual content must be checked against official or authoritative sources before publication.
Clarvia does not intend to collect personal bereavement case data in phase one.
Do not add:
- identity documents,
- death certificates,
- family details,
- addresses,
- financial information,
- medical information,
- private correspondence,
- or personal case descriptions
to public repositories.
If an error is found, Clarvia should:
- identify the affected object;
- check the relevant official source;
- update the data or mark it as stale;
- record the correction where appropriate;
- rerun validation;
- update generated outputs if needed.
Corrections should prioritize user safety, source accuracy, and transparency.