Generative AI in BFSI Industry: Document Analytics Platform
Developing a Generative AI-based document analytics platform to extract pertinent entities from a variety of file formats, such as .pdf, .xls, and .doc, originating from multiple sources. The extracted data is visible through user dashboard and chat window, powered by OpenAI’s GPT-4 model.
THE CUSTOMER
BUSINESS GOALS
The customer would like to perform business analytics over various financial documents that they intake on a regular basis as an audit firm. They asked ThirdEye to develop a platform which can process and extract data from large volume of files in different formats like .pdf, .doc, .xlsx etc.
They wanted to leverage Artificial Intelligence (AI), Process Automation, Machine Learning (ML), Natural Language Processing (NLP) and ETL processes for this project to enable their firm to make strategic business decisions while reducing manual labor.
THE SOLUTION
ThirdEye proposes to commission a managed solution named “Optira” that will have all required functionalities and will scale over time as per business traction.
The proposed platform is designed to be extensible and scalable, to meet ever-changing business needs with the capability to extract pertinent entities from a variety of file formats, such as .pdf, .xls, and .doc, originating from multiple sources. The platform also has the feature to activate ChatGPT for accessing the extracted data.
ThirdEye has deployed the Generative AI based document processing platform Optira to meet the business goals. The platform has all the required features which the customer has asked for. Along with the primary features like data ingestion from multiple sources, automating data processing for various file types, it has been quipped with PowerBI dashboard to visualize and analyze the data and ChatGPT interface to ask questions and get extracted answers in normal English.
The Optira Solution would be/have:
- Designed to be extensible and scalable, to meet ever-changing business needs.
- Developed with the aim of minimizing implementation & operational costs.
- Prioritized to achieve highest ROI as quickly as possible.
- Adaptable to cater to the requirements of diverse industry needs.
- Available on a pay-per-use basis.
- Can be re-sold as a white-labeled solution.
- Offers a more cost-effective solution, as its charges can be classified as operational expenditure rather than capital expenditure.
- Available support services as per terms specified by the customer.
- Capability to extract pertinent entities from a variety of file formats, such as .pdf, .xls, and .doc, originating from multiple sources.
- Capable of extracting “tables/forms” and various “keys & values” from the source data.
- Extracted data stored in a structure format like SQL or NoSQL table, or saved in other format like .pdf, .xls, .doc.
- Data merge or aggregated into a common table or format.
- End-to-end flow automated with custom web UI application.
- PowerBI based data visualizations and analytics.
- Inbuilt OpenAI-ChatGPT capabilities for consumers to query in simple english and get back detailed and specific textual responses, including graphs and charts.
Technologies Incorporated:
- Azure Blob Storage
- Azure Event Hub
- Azure Form Recognizer
- Azure Data Factory
- Azure Cosmos DB
- Azure Machine Learning
- PowerBI
- OpenAI’s GPT-4 Model
- Azure Computer Vision
- Azure Cognitive Services
- Custom Classification Models
- Custom Web UI Application
VALUE CREATED
ThirdEye has deployed the platform for customer’s internal use, successfully running on stakeholders’ systems as per business requirements. The generative AI-powered platform is delivering the desired results by processing large volume of files. The customer is very happy with the result and requested ThirdEye to work further on the platform to make it ready as managed solutions.
Transforming Enterprises with
Data & AI Services & Solutions.
ThirdEye delivers Data and AI services & solutions for enterprises worldwide by
leveraging state-of-the-art Data & AI technologies.
