In addition to Reuters Dealing 3000 conversations, the DealHub Conversation Analyser has the ability to extract and create tickets from other unstructured text based sources such as secure messaging systems, e-mail and their attachments. Tickets are extracted by applying proprietary parsing algorithms to provide a solution which is scalable, prevents duplicates and is extensible to cater for new trading venues in the future.
Peter Kriskinans, Managing Director of Option Computers commented âMultiple deals can be found within a conversation and the Conversation Analyser can accurately analyse conversations to extract multiple tickets. The trades which are parsed from the conversation text can also be formatted and sent via STP to matching or risk systems which enables the bank to verify trades much earlier than in the current workflow.â
The DealHub Conversation Analyser has the ability to setup highly configurable parsing patterns which provides the IT department and end user with the ability to customise different sources, instrument types and workflows as well as to define word lists and aliases for enhanced pattern recognition.
Parsing can take place on the traderâs desktop in real-time, or in middle and back office âpost-tradeâ as part of downstream workflows. All incoming conversations and chats are also stored in a 10 year SOX compliant archive, creating the mandatory compliance and audit record of trading activity that facilitates investigations.
âScoring thresholds can be set which enable financial institutions to determine the validity of the transactions which are created. The ability to associate commonly misspelt terms with existing library terminology allows super-users to control additions to the parsing libraries on a global basis, whilst still allowing local users to align parsing against local trader terms. This allows the system to be continuously extended to cover new instrument types.â said Kriskinans