The Smart Engines server SDK family allows you to create high-performance in-house recognition services that efficiently and securely solve the tasks of optical text recognition, recognition and processing of ID documents, as well as business documents. The product line with three SDKs (Smart Code Engine, Smart ID Engine and Smart Document Engine) allows you to create modern recognition services to meet the needs of medium sized and large enterprises. Besides, full optimization and multi-core processor support allow you to upscale solutions, achieving the recognition system performance that the customer needs. The products support automatic recognition of scans and photos: bank cards of 21 payment systems; 1D and 2D barcodes; passports, driver’s licenses and national ID cards of 210 jurisdictions; as well as questionnaires, application forms, financial, payment, trade and accounting documents in 102 languages worldwide.
All Server OCR SDKs are built on a single proprietary computer vision platform of Smart Engines, can be seamlessly integrated and can be extended with custom functions within simple and effective APIs. This allows developers of document processing services to use the functionality they need, and optimally solve their business tasks with no risks of image transmission to third-party services. Constant intensive load testing, careful optimization of RAM usage and computational resources consumption, as well as proprietary optical recognition of industrial quality create high performance and high security enterprise solutions.
With the help of the Smart Engines server product line, you can create enterprise services or add to an existing data exchange bus: text data recognition of bank cards and barcodes for financial acquiring processes and KYC; scanning and verification of passports, national and international ID cards to provide remote service for customers and employees, banking, financial and insurance retail, or telecommunication services on-boarding; statutory document recognition for KYC/AML or legal entity audit, questionnaires and official forms scanning for G2C (Government-to-Citizen) and G2B (Government-to-Business) process automation; document recognition for integrated accounting and management automation. Our products perform all calculations on-premises, do not transfer data to third-party servers and do not use manual data entry.
It is important to highlight that the software does NOT transfer your customers’ personal data for processing to third-party services and/or third parties for manual data entry, does NOT save data: all processing is carried out in the device’s local RAM, network connection is NOT required for recognition. All the solutions are HIPAA, GDPR, CCPA-compliant.
Green AI-based computer vision technologies developed by our scientists and engineers allow us to minimize the carbon footprint of AI technologies. The GreenOCR® optical text recognition technology created by our scientists and engineers allows us to achieve the highest recognition quality in 102 languages on any device, in real time, the only computing resources required are those of the service itself. Careful algorithm optimization and implementation make it possible to use energy-efficient server complexes for optical recognition. The recognition performance on a modern 32-core processor reaches 45 ID documents per second.
The optical recognition product line is delivered as an SDK with documented OCR API and the necessary code samples. The product line supports all major modern computing architectures, is specially optimized for: x86_64, ARM, MIPS, and runs under Windows operating systems (officially supported versions), MacOS, and Linux-based systems.
In our company, the Quality is not only recognition precision, but also efficiency, reliability, and predictability of all algorithms. In order to achieve uncompromisingly high precision and document recognition performance in services, our scientists and engineers constantly develop the product architecture, new algorithms and optimize their implementations for server computing architectures.
Environment Friendly —
Green AI in DNA
When creating our products, as a priority, we take into account the potential impact of the technologies and algorithms on the environment. To minimize energy consumption in the recognition process, we created special energy-efficient architectures of low-bit extra-compact neural networks, a special training approach does not require much computational power. Thanks to them, we released the first industrial-class OCR in the Green AI ideology, which helps our clients to reduce carbon footprint and pursue the goals of the United Nations Sustainable Development Program.
In our developments we train AI to help humans, not to train humans help the AI. For this purpose, we created a recognition system that works in a wide range of angles, scales, lighting conditions (bright light, glare, twilight). The products have an auto-selection function for the document type, which performs document streaming processing in a fully automatic mode. To achieve high performance, our products do not require specialized servers with computing accelerators, thanks to software optimization, the products provide the highest performance on ordinary servers.
An essential aspect of our SDKs is compliance with ethical and legal norms. All SDKs are created using synthetic data and free of privacy violation or ethically suspicious technological solutions and approaches. We test our products for compliance with legislation in the personal data processing, for hidden functionality and comply with the requirements of HIPAA, GDPR, CCPA.
Smart Code Engine is a server SDK for bank card data scanning, 1D and 2D barcodes (QR codes, AZTEC, PDF417, DataMatrix, EAN, UPC, etc.), machine-readable zone (MRZ) of passport and visa documents (MRTD, eMRTD). In addition, it is possible to scan price tags, product labels, VIN numbers, serial numbers of parts and assemblies. Server API provides the ability to create data recognition services from single images (scans, photos) and in a real-time video stream of any performance. Smart Code Engine is intentionally optimized for efficient and fast operation within the application in the service development process.
Smart ID Engine is a server SDK which scans and verifies passports, national ID cards, residence permits, work permits, visas, driver licenses, vehicle registration certificate, birth, marriage, divorce, death certificates and other documents related to identification and property rights. The ID scanning server SDK provides secure API which creates services that solve the tasks of remote identification and verification by data optical recognition from single images (scans, photos) and in a real-time video stream. Our special fraud detection algorithms detect attacks and protect against fraudsters. Smart ID Engine is specially optimized for fast and precise recognition on devices with limited computing power.
Smart Document Engine is a server SDK with precise and secure API for scanning forms and extracting data from documents. The SDK is designed to automate the workflow, optimize document entry and intelligent document processing. The server SDK quickly and securely performs optical text recognition of business, statutory, accounting, tax, notarial, legal, insurance and banking documents, as well as standard questionnaires and forms of strict accountability. The simple API allows you to quickly integrate document scanning into any services and enterprise service buses.
NEC de Colombia
Sum&Substance, a global KYC/AML service provider, use Smart ID Engine for IDs scanning
Please fill out the form to get more information about the products,
pricing and trial SDK for Android, iOS, Linux, Windows.