It’s “Early Days” for NLP. The Opportunities Are Many

It’s “Early Days” for NLP. The Opportunities Are Many

Last Week I Participated on an AI-focused Panel at the SS&C Deliver Conference

The panel – “Practical Applications for AI in Investment Operations & Accounting” – was especially enjoyable because the audience engaged. While much has changed since I last worked in an Investment Operations capacity in 1998, much remains the same – meaning there is abundant opportunity for further automation and operational efficiency gains – only this time AI-related technologies will be the driving force.

We Are Generally Bullish About NLP Across Investment Ops and Similar Use Cases

During the brief Q&A session the panel fielded at least one if not two questions about Natural Language Processing (“NLP”). I was anticipating NLP-related questions because Scott Kurland of SS&C had walked me through an impromptu 20 minute demo of Singularity (SS&C’s Artificial Intelligence-infused investment operations platform), earlier that morning where I had the opportunity to experience the new NLP-based features along with Singularity’s sleek, enhanced User Interface.   

One of the NLP-related questions I remember distinctly as it was related to the manual nature of processing bank loans in the backoffice. The woman who asked the question essentially said “I’m manually entering bank loan data, can you help automate the process?” The answer of course was “YES”. For context, bank loans were among the fixed income security types that required the most manual intervention dating back to my time in Putnam’s backoffice. Today, Singularity has the ability to process securities such as bank loans in automated fashion. In the event that a particular bank loan gets flagged for lacking certain data elements, rather than the user having to manually enter the missing elements, Singularity leverages NLP and optical character recognition (“OCR”) to scan the source document, extract the relevant information and auto-populate the missing data fields where the user may review those fields prior to approving (data fields are marked for accuracy by the system with assigned confidence levels). Any corrections the user may make are “remembered” by Singularity, thus the platform is perpetually learning. I fully expect that Singularity’s new NLP-based capabilities will drive incremental revenue.

Similar Prospective NLP/OCR Use Cases

Similar use cases include any process where information from a source document is manually entered into a system of record. These examples include insurance Policy Administration systems, general ledger/ Corporate Treasury systems, real estate transaction systems, customer on-boarding, deal sourcing platforms and the list goes on. Stay tuned to TEK2day as we track NLP progress in the coming years.    


Advertisements