Computer systems cannot operate upon data embedded in human-generated text because it is free-formatted and lacks a predictable structure. In the absence of text interpretation software, organisations must manually extract data from free-formatted text and re-key the extracted data into the system that will ultimately act upon it.
Manual data extraction and re-keying is a slow, costly, error-prone and labour intensive task that creates processing bottlenecks and data quality issues. The recent explosion of free-form electronic communication such as email, text messages, chat and social media means that organisations will increasingly need to add automated text interpretation capabilities to their data processing infrastructure.
SyntelliRead is focused exclusively on developing text interpretation software for the financial industry. Our applications accurately and intelligently interpret, extract, and structure data and information embedded in free-formatted text in order to make it accessible to other systems.
By functioning as a human-to-machine translator, our solutions add value by dramatically reducing the manual labour, time, cost, and error rates associated with processing unstructured information, thereby improving organisational efficiency, consistency, quality, and responsiveness.
Anti-money laundering systems need to identify transactions occurring in FATF non-cooperative countries and territories (NCCTs), countries subject to OFAC sanctions, high-intensity drug trafficking areas (HIDTA), high intensity financial crime areas (HIFCA), and other high-risk jurisdictions. But the ability of AML systems to detect such transactions is severely limited because the information about where transactions are taking place is specified in free-formatted party address fields.
The city, province / state, and country names embedded within those fields are inaccessible to AML systems in their raw form and must be extracted. However, extraction of these names is difficult because party addresses are often incomplete, ambiguous, and prone to spelling variations and misspellings. Extraction errors compromise an AML system's ability to accurately perform geographic-based transaction monitoring.
SyntelliRead Location Extractor accurately extracts and standardises city, province, and country names from free-formatted party addresses, thus providing AML systems with the geographic data needed to identify high-risk transactions. SyntelliRead Location Extractor is compatible with all AML systems, can interpret any number of addresses in any record format, and does not require any third-party software or database.
Some of the world's largest banks use SyntelliRead software as a front-end to their exception management systems to facilitate the processing of free-formatted inbound SWIFT, CHIPS, FED and email investigation messages.
SyntelliRead Investigations Parser extracts customer, payment and case information from inbound investigation messages, thus enabling workflow engines to automate tasks such as opening new cases, attaching messages to existing cases, creating links between cases and related transactions, routing messages to their appropriate workflows, and generating automated responses.
By using SyntelliRead Investigations Parser as a message pre-processor, our customers dramatically reduce the time, cost and errors associated with handling free-formatted messages by eliminating manual data re-keying.
Our text interpretation software has been used by some of the world's largest financial institutions for over a decade. Unlike simplistic keyword matching technologies, our solutions combine powerful, noise-tolerant pattern matching with sophisticated reasoning mechanisms specifically designed to deal with the inconsistency, ambiguity, and incompleteness that is common in free-formatted text.
Our core technology is designed to handle text in any format, regarding any subject in any industry, and can be applied to a variety of text-based applications across the entire enterprise, including inbound message processing, outbound message screening, customer comment gathering from social media, and data cleansing.
As a result of our 25+ years of commercial experience in natural language processing (NLP) and text interpretation technology research and development, our applications are accurate, fast, fault tolerant, reliable, easy to install and easy to maintain.