Three gaps in China-source monitoring
Teams that follow Chinese government activity face connected problems that most existing tools only partially address.
Source velocity
Ministries, the State Council, the NPC, and party signal channels publish in overlapping cadences. Manual monitoring is slow and incomplete.
Evidence fragility
When a source page updates, the previous version disappears. Without version capture, analysts cannot cite what a page said on a specific date.
Language & normalization
Source content is in Chinese. Entity names appear in multiple forms. Teams working in Japanese or English need consistent identity resolution across languages.
From collection to structured research output
XINAPI closes all three gaps in a single, auditable pipeline.
Continuous source monitoring
Scheduled ingestion across 6 priority source layers. Every fetch is versioned and hashed. Automated cron runs every 6 hours.
Structured diff & version tracking
Line-level diffs between source versions. Raw text diff and block-level comparison. Every diff references explicit version pairs.
Multilingual entity resolution
Canonical entity IDs with native, Japanese, and English names. Language-independent identity designed for ja/en/zh from day one.
Source-linked claims & relations
Claims and relations are extracted with traceability to the specific source version they came from. No detached summaries.
Human review workflow
Claims and relations go through review states: pending, accepted, rejected, revised. Only accepted content flows into outputs.
Report assembly & export
Reports compile entities, claims, relations, and source-linked evidence into structured packages ready for downstream use.
23 sources across 6 priority layers
Authoritative, research-relevant, and traceable. All Layer 1, 2, and 3 sources are actively ingested in production.
Seven-step analysis pipeline
From source registration to structured export, every step maintains traceability back to the original source.
One pipeline, three surfaces
Separate surfaces for separate audiences, each with distinct access controls and purpose-built interfaces.
XINAPI API
Programmatic access to source collection, entity resolution, diff, and report endpoints. Public v1 endpoints with rate limiting. Admin routes behind Zero Trust.
- Automated source ingestion and version tracking
- Entity search and resolution across languages
- Diff queries between source versions
- Report creation and export via endpoints
XINAPI Platform
Browser-based workspace for source browsing, diff viewing, entity search, claim review, and report management. Multilingual UI in Japanese, English, and Chinese.
- Browse and inspect monitored source documents
- Review diffs and track change history
- Accept or reject extracted claims and relations
- Build source-linked reports for export
XINAPI Infrastructure
Internal control surface with health monitoring, ingestion diagnostics, audit logs, and deployment verification. Access restricted via Zero Trust.
- Monitor ingestion pipeline health and status
- Inspect audit logs and delivery records
- Verify deployment state and system diagnostics
- Manage retention and policy controls
Live on production today
Verifiable via public API endpoints. Not staging, not planned — running now.