Generation of ToC for Tax Transparency Reports with dynamic linking
Our client, a global financial professional services firm, sought to optimize its document management processes by efficiently extracting and templating the Table of Contents (ToC) data while ensuring accuracy and alignment with industry standards
Outcome
10%
Efficiency gain allowed the teams to focus on higher-value tasks.
20%
Increased proactive proposals, giving them competitive advantage.
4/5
Customer satisfaction scores– context specific relevant information to the business users.
Challenge
Client’s analyst teams were spending a lot of time with a lot of error rate due to manual processing while extracting the Table of Content (ToC) data from Tax Transparency Reports (TTRs) and creating standardized templates
Our Solution
- Built a solution, leveraging advanced Generative AI technologies
- Our solution incorporates dynamic linking capabilities, enabling navigation between the extracted TOC and relevant sections within TTR documents for easier summarization/ information retrieval mechanisms.
Using the open-source LLM, Llama2 architecture, we ensured secure storage and access to TTR documents locally within the company’s premises, prioritizing data privacy and security.
Ensemble of LLM recognition, OCR Deep learning models and a rule-based approach for a highly effective retrieval of table of contents.