Ephesoft
Ephesoft, now Tungsten Transact following its 2022 acquisition by Tungsten Automation, provides intelligent document capture and processing solutions combining OCR, machine learning, and workflow automation.
Overview
Ephesoft, founded in 2010, developed cloud-native intelligent document processing technology for converting unstructured documents into structured data. Tungsten Automation (formerly Kofax) acquired Ephesoft in August 2022 to integrate its document processing capabilities into Tungsten's broader intelligent automation platform. At acquisition, Ephesoft reported year-over-year growth exceeding 30%.
The platform, now branded as Tungsten Transact, serves organizations across financial services, healthcare, government, insurance, and manufacturing sectors. The software uses supervised and unsupervised machine learning to automatically classify documents, extract data, and route information to business systems with minimal template configuration required.
Key Features
- Intelligent Document Capture: Multi-engine OCR processing digital and paper documents across 100+ languages
- ML-Based Classification: Automatic document categorization using supervised and unsupervised learning models
- Data Extraction: Field-level information capture from structured and unstructured documents
- Handwriting Recognition: ICR technology for processing handwritten forms and documents
- Cloud-Native Architecture: Scalable deployment in cloud environments with hybrid and on-premise options
- Minimal Training Requirements: Document recognition from as few as one sample document
- Multi-Channel Capture: Processing from scanners, email, web portals, and mobile devices
- ERP/ECM Integration: Connections to major enterprise content and business systems
Use Cases
Mortgage Loan Processing
Financial institutions process loan applications and supporting documents including tax returns, bank statements, and property appraisals. The system classifies over 50 document types common in mortgage packages, extracts borrower information and property details, validates data against business rules, and integrates with loan origination systems for automated applicant record population.
Healthcare Records Management
Healthcare providers digitize patient medical records and administrative documents through automated capture of registration forms, insurance cards, referrals, lab results, and clinical notes. Machine learning algorithms classify documents and extract patient demographics, insurance details, medical record numbers, and diagnostic codes for integration with EHR and practice management systems.
Technical Specifications
| Feature | Specification |
|---|---|
| Deployment Options | Cloud, Hybrid, On-premises |
| OCR Technologies | Multiple engines for optimal recognition |
| Machine Learning | Supervised and unsupervised learning models |
| Extraction Accuracy | Up to 99.9% for structured documents (claimed) |
| Processing Speed | Up to 95% faster processing (claimed) |
| Supported Languages | 100+ languages |
| Integration Methods | REST APIs, CMIS, Web Services |
| Document Types | Invoices, contracts, forms, correspondence, handwritten documents |
| Security Features | Encryption, role-based access, audit trails |
| Pricing | Consumption-based starting at 120,000 pages/year |
| Pricing Tiers | Standard, Professional, Enterprise |
Getting Started
- Requirements Assessment: Analyze document processing needs
- Solution Design: Configure document types and extraction rules
- Implementation: Deploy with integration to existing systems
- Model Training: Train machine learning models on document samples
- Optimization: Continuous improvement of extraction accuracy
Resources
Company Information
Founded: 2010
Acquisition: Tungsten Automation (August 2022, undisclosed amount)
Former Name: Ephesoft
Current Branding: Tungsten Transact
Growth at Acquisition: 30%+ year-over-year
Industries: Financial services, healthcare, government, insurance, manufacturing
