Intelligence OCR — Extracting IOCs from Images
Multi-language OCR over images, screenshots and PDFs, with NER (Named Entity Recognition) specialized in IOCs: domains, IPs, hashes, BTC/ETH addresses and Brazilian personal tax IDs (CPF).
Overview
Extracted entities are normalized and indexed, becoming available for search, graph correlation and alerting.
Capabilities
- 180+ supported languages
- NER specialized in IOCs
- Full-text indexing
- Confidence scoring
- Bounding box export
Use Cases
- Extracting IOCs from forum screenshots
- Indexing report PDFs
- Analysing Telegram screen captures
- OSINT on images
Integrations
- Tesseract, Google Vision, Azure CV
- STIX 2.1 export
SLA & Guarantees
Immediate OCR per page