Product Querier for multi-platform documentation & support for an American Financial Services Company
Overview.
The client develops hundreds of financial products for internal and external consumption. These products are multi-platform & multi-technology based and have a large set of documentation around them, running into hundreds in many cases. Enhancements and support to these products become very tedious & time taking due to the large set of documentation and at times minimal knowledge of the product. The objective of the solution “Product Querier" using GenAI is to provide users with an efficient and accurate way to search and retrieve information from a vast collection of documents.
Hundreds of products where each product has 100+ of documents in multiple formats.
These are long-term projects with ongoing enhancements, support and cross-product needs.
Frequent changes in development team resulting in ongoing learning curve.
Delays in enhancements due to time taken in understanding previous work impacting thousands of customers/financial institutions utilizing these products.
Information is available in a variety of forms such as, textual, flow charts, graphs, excel sheet, presentations, video etc.
Using manual search methods proved to be inefficient and inaccurate.
Solution.
The solution utilizes GenAI, an advanced artificial intelligence technology, to analyze and understand the content of documents.
Solution is developed using Autonomous agent which can operate independently and interact in natural way.
All the documents are loaded to vector database from Knowledge base.
It then enables users to enter specific queries and retrieves relevant information from the documents in a quick and precise manner.
A chat bot interface is provided for SME and sales team to interact with the agent and get response to queries. Response sources are also provided along with the query response. Users can further go onto the specific page of document for more information on the user interface.
API based integrated with the upstream and downstream application also supported.
The Impact.
Revenue Optimization: Up to 20% faster product go-live
Cost optimization: Up to 10% lower cost per project
Operational efficiency: Documentation Retrieval faster by up to 40%
Search Accuracy: Initial Accuracy up ~75% Accuracy post 3M up ~88%