Global PEP database including Nigerian federal/state officials, NASS members, military, judiciary, and family members. Tiered PEP classification (Tier 1–3). Updated from official government sources.

Covers English, Yoruba, Hausa, Igbo, Pidgin, French, Arabic. Sources: Punch, Vanguard, ThisDay, Guardian NG, Premium Times, Daily Trust, social media. NLP relevance scoring and entity disambiguation.

Searches Nigerian federal and state court dockets, bankruptcy filings, judgments, and liens. Returns structured case data: parties, suit number, court, filing date, status, and disposition. Covers Federal High Courts, State High Courts, NJC records, and CAC winding-up petitions.

Standardizes Nigerian addresses (street, LGA, state). Cross-references against known virtual office database (Regus, Spaces, coworking hubs). Validates physical presence via Google Maps API and satellite imagery. Flags PO boxes, mail drops, and coworking fronts.

Discovers and correlates social media accounts (LinkedIn, X/Twitter, Facebook, Instagram), website registrations (WHOIS), review profiles (Google Business, Jumia), and online mentions. Maps an entity's digital presence to verify identity claims and surface hidden connections between accounts.

Cross-references device identifiers, browser fingerprints, and behavioral patterns against known fraud networks. Detects emulators (Bluestacks, Genymotion), rooted/jailbroken devices, device ID spoofing, and SIM swap events. Returns device risk score, fraud network membership, and behavioral anomaly flags.

Resolves IP addresses to geolocation (city, state, country), ISP, and hosting provider. Detects proxies, VPNs, Tor exit nodes, and residential proxy networks with precision scoring based on threat intelligence feeds. Returns risk score, network type, and historical abuse reports.

Maps beneficial ownership structures through multiple corporate layers. Pulls from 50+ jurisdictional registries, Panama Papers, OpenCorporates, CAC, and leaked corporate data. Returns ownership percentages, director histories, and connected entity graphs.

Nigeria: CAC (RC Number, CAMA filings, annual returns). North America: SEC EDGAR, Canadian SEDAR, Mexican RNIE. Europe: Companies House UK, German Handelsregister, French SIRENE. Asia-Pacific: ASIC Australia, Singapore ACRA. All returns structured with officer histories and filing status.

Nigeria UK US Canada UAE +47 more

AI-powered headless browser agent that autonomously navigates the web to research entities, gather evidence, and extract structured intelligence. Follows links, handles JavaScript-rendered pages, and returns structured findings with source citations. Used for OSINT, regulatory lookups, and company verification.

Pulls structured data from any target webpage — including dynamic, JavaScript-rendered sites. Normalizes extracted content into typed data fields: names, addresses, dates, amounts, entity relationships. Used to extract data from court portals, CAC filing pages, news sites, and social profiles.

Multimodal document processing. Downloads and renders PDFs, images, and scanned documents. Extracts key fields, tables, signatures, and stamps. Returns structured data with visual annotations. Used for CAC certificates, court orders, bank statements, and contracts. Powered by AWS Textract + vision models.

High-accuracy OCR engine for scanned Nigerian documents. Handles poor scan quality, handwritten annotations, stamps, and watermarks. Extracts fields by document type (passport, NIN, utility bill, certificate of incorporation). Returns typed JSON with field confidence scores and bounding box coordinates.

AI agent that can operate computer interfaces — including legacy internal bank systems, core banking platforms (Finacle, T24, Flexcube), and apps without APIs. Takes screenshots, clicks, types, and extracts data from any application. Enables AML investigation without API integration. Sandboxed and audited.

Resolves entity identities across fragmented data — matching partial names, typos, aliases, and transliterations. Detects known fraud patterns (layering, structuring, round-tripping) at entity level across multiple accounts and institutions. Discovers hidden connections using graph analysis on shared identifiers (phone, address, device, email).

Given any identifier (BVN, NIN, phone, email, device ID, address, IP), traces all connected accounts, entities, and users across your institution's data and TRAMCA's network graph. Returns a visual entity map with connection strength scores. Critical for uncovering ring fraud and shell company networks.

Compiles the complete investigation into a structured, defensible output package. Includes: risk decision with confidence score, full chain of evidence (each tool's input/output), reasoning traces, source links, analyst notes, and SAR narrative. Exported as PDF or JSON for CBN submission, internal records, and goAML filing.

Validates government-issued identity documents including NIN slips, International Passports, Driver's Licenses, Voter's Cards, and National IDs. Extracts identity fields (name, DOB, ID number, expiry), checks document security features (holograms, machine-readable zones, chip data), and cross-references against NIMC/FRSC/INEC issuing authority format templates for each document type.

Analyzes documents for signs of manipulation including pixel-level tampering (clone-detect, splicing, noise inconsistency), font inconsistencies (character spacing, baseline drift, mixed typefaces), metadata anomalies (creation date, software fingerprints, GPS data), and template mismatches against known genuine document layouts from Nigerian issuing authorities.

Analyzes image artifacts, screen re-capture patterns, compression signatures, and texture inconsistencies to distinguish genuine camera-captured documents from screenshots, scans of scans, or synthetic images. Detects moiré patterns, LCD sub-pixel patterns, and jpeg re-compression artifacts that indicate a screenshot was taken of a screen rather than a physical document.

Flags deepfake selfies and AI-generated document images using model-fingerprint signals (GAN artifacts, diffusion model signatures), rendering artifacts (unnatural lighting, blurred hairlines, inconsistent skin texture), and consistency checks across face, text, and layout regions. Detects Stable Diffusion, Midjourney, DALL-E, and face-swap outputs. Critical for preventing synthetic identity fraud.

Evaluates handwriting samples across multiple documents to identify likely forgeries and inconsistencies. Analyzes stroke patterns (curvature, speed signatures, pen lifts), pressure mapping (ink density, line weight variation), and character-level matching (individual letterforms, proportions, baseline consistency). Used to verify signatures across contracts, board resolutions, and identity declarations.