Trading records reconciliation for investment firm
95%
straight through processing
2000+
statement structures processed daily
90%
labor-savings
Client
Top Asian Investment Firm
Goal
Reconcile over two thousand types of transaction statements daily
CHALlENGE Overview
Investment firm dealing with trading records from different parties with various file formats and layouts resulting in over 2000 types of data structures.
- The firm was using a variety of solutions to extract information from the differing data structures, but the growing variety of data made the process inefficient and difficult to manage.
WHAT WE DID
- AI model built within days to process various layouts, formats, and data types.
- Integrated solution to transform data from unstructured to structured and feeding that data to JSON format for use in several systems of record.
- Fully automating trade confirmation statement reconciliation, using AI and ML to provide the highest level of accuracy.
IMPACT
- $1.7 Million+ annual savings
- 90% reduction in manual efforts which were prone to errors
- 95% straight-through processing with automated human in the loop processing
- 2000+ Data types - Thousands of statement structures and formats processed daily
