Know who crosses before they board.
AI-powered risk intelligence for every passenger on every flight — PNRGOV ingestion, eight threat categories, 38+ expert rules, and full case management from first alert to detention. Deployed in under 30 days.
BorderSentinel ingests PNRGOV and API transmissions in real time via the built-in PNR ingestion service and ADS-B flight tracker. Each inbound flight surfaces its risk count before the aircraft is wheels-down.
BorderSentinel's rule engine is organised into eight specialist threat domains. Each domain carries its own weighted rules, written by customs and border security practitioners, not generic ML engineers.
Overstay risk, visa-class mismatch, transit-without-visa abuse, known irregular migration corridors, and prior removal history.
8 rulesHigh-risk drug-source routing, mule profile indicators, frequent short-haul trips to production countries, and courier booking patterns.
5 rulesUndeclared goods corridors, routing through low-duty transit hubs, and group-travel commodity smuggling patterns.
4 rulesHuman trafficking indicators: group same-gender travel, third-party ticket purchase, no return booking, economic vulnerability corridor routing.
6 rulesTravel to conflict zones, watchlist proximity, financing patterns, and known radicalization corridor itineraries.
4 rulesMoney laundering courier profiles, hawala corridor routing, bulk-cash movement risk, and high-frequency short-trip financial hub patterns.
4 rulesMismatched identity across bookings, document-country inconsistency, name variant clusters, and prior-known false identity travel.
4 rulesAnomalous trip frequency, rapid country cycling, itinerary contradiction, unexplained multi-leg routes through high-risk transits.
27+ HISTORY rulesA traveller who changes their passport doesn't reset their risk profile. BorderSentinel links identities across all three resolution tiers and builds a unified cross-travel record.
Exact match on passport number and issuing country. Highest confidence — the same document used again.
Full name + date of birth exact match. Catches passport renewals and deliberate document changes while preserving the same biographical identity.
Levenshtein distance scoring on name variants — transliteration differences, aliases, and misspellings — so slightly altered names don't escape the match.
Every risk score in BorderSentinel is fully decomposable. Analysts see exactly which rules fired, the score contribution of each, and the data that triggered them — evidential grade from day one.
Administrators can author new rules directly in the UI — define trigger conditions, set score weights, choose the threat category, and publish. No code deployment required. Rules go live on the next PNRGOV ingestion cycle.
Cyber-scam centres exploit thousands of trafficked workers across South-East Asia. BorderSentinel includes specialist EXP-SCM rules grounded in IOM and UNODC research to identify victims before departure — not after arrival.
Why pre-departure? Victims are frequently held under duress and may not self-identify at the border. Catching the trafficking pattern in the PNR — third-party tickets, group same-gender same-age booking, economic-vulnerability origin — enables interception before the victim ever leaves their home country.
BorderSentinel carries a complete case-management workflow so border officers can act on intelligence without switching systems. Every action is logged, timestamped, and audit-ready.
Automatic legal-hold timer starts on case creation. Configurable maximum-detention thresholds with escalation alerts ensure compliance with domestic detention limits.
Prosecution decisions require sign-off from two authorised officers at Supervisor or Investigator level. No single officer can unilaterally escalate to prosecution.
Every piece of evidence — documents, digital files, statements — is logged with handler, timestamp, and hash. Chain is unbroken from seizure to court submission.
Real-time de-conflict across all open cases and concurrent investigations. Prevents two officers unknowingly working the same subject or disrupting an ongoing surveillance operation.
Structured interview templates with configurable question sets by threat category. Responses captured in structured format for rule-feedback loops and analytical re-use.
Configurable notification rules push alerts to mobile devices, email, and agency systems. Escalation paths defined per threat category with SLA tracking.
From flight-level overviews to seat-level cabin maps and passenger relationship networks — BorderSentinel gives officers the visual context to make fast, informed decisions.
Every BorderSentinel user operates within a strictly scoped role. Sensitive case actions require elevation — no single officer has unbounded access.
BorderSentinel is a microservices platform. Each service owns a single responsibility and exposes a clean API — deploy the whole stack or integrate individual services into your existing infrastructure.
Real-time aircraft position and ETA ingestion from ADS-B feeds — triggers pre-arrival processing at configurable lead times.
Accepts IATA PNRGOV format and direct airline API. Parses, validates and normalises passenger data for the risk engine.
Orchestrates identity resolution, rule evaluation and score aggregation. Stateless, horizontally scalable — handles peak inbound waves without queueing.
Weighted rule aggregation, band assignment (CRITICAL / HIGH / MEDIUM / CLEAR), and score history storage for trend analytics.
CRUD API for rule authoring, versioning, activation and rollback. Powers the no-code Rule Builder UI for operational teams.
Manages officer examination workflows, structured interview capture, and result recording — feeds outcomes back to the scoring model.
Push, email, and webhook dispatch. Configurable alert thresholds per threat category with SLA tracking and escalation chains.
Full case lifecycle — detention clock, dual-auth prosecution, exhibit chain-of-custody, deconfliction, and court-bundle export.
Single authenticated entry point for web UI and external agency API consumers. Role-based request filtering, rate limiting, and audit logging.
Talk to our team about your inbound volume, threat priorities, and integration requirements. We'll scope a pilot that goes live in 30 days.