The Problem That Used to be
Research organizations serve as the backbone of policy development, analyzing critical data to suggest systemic improvements. TrustBridge, one such leading research organization, actively engages in studying regulatory enforcement actions and recommending reforms. In the course of its mission, TrustBridge frequently encounters the challenge of extracting valuable insights from extensive and disorganized regulatory order datasets. Faced with this complexity, the organization found itself at a crossroads, where traditional manual processes could no longer keep pace with its ambitious goals.
In its mission, TrustBridge is tasked with identifying patterns and insights from orders issued by various regulators and regulatory tribunals. These documents are essential for understanding regulatory trends and proposing necessary reforms. Each project demands the identification of specific data points, known as “indicators,” which facilitate meaningful analysis of the orders.
These indicators vary across projects. For instance, one project - on orders of the Securities Appellate Tribunal - required TrustBridge to extract 80 unique indicators from a dataset of over 500 tribunal orders. The extracted information was then collated into detailed reports, forming the basis for implementable recommendations.
Despite the clear objectives, the process is far from straightforward. Relying heavily on manual data extraction, TrustBridge researchers meticulously combed through hundreds of orders, identifying and recording data points in line with the indicators. While this method ensured that the insights were grounded in a thorough understanding of each document, it also introduced several challenges.
Resource Intensive: The manual extraction process demands considerable time and personnel. Large teams are required to process hundreds of documents based on specific indicators, which can strain the organization’s resources, especially for larger projects.
Subjectivity & Human Error: Despite rigorous training and established procedures, interpreting complex indicators remains subjective. Researchers often encounter nuances in the text that can lead to variations in interpretation, compromising the uniformity of results. The sheer volume of data increases the likelihood of human error, while repetitive tasks leave little room for deeper analytical work.
The Lucio Solution
TrustBridge partnered with Lucio to reimagine their approach to studying regulatory enforcement action. Lucio’s AI-driven tools allowed TrustBridge to automate their data extraction processes, ensuring accuracy, consistency, and scalability.
Lucio worked closely with TrustBridge to understand the unique requirements of their projects. TrustBridge is now able to define indicators for each of their projects. Lucio instantly reviews the orders, and extracts all details required to complete indicators. Lucio’s AI review accounts for a substantive and contextual analysis of each document. Moreover, TrustBridge researchers can easily audit Lucio’s output through clickable hyperlinks to source pages.
Finally, Lucio converts the responses into strict output types desired by TrustBridge researchers. Thus, Lucio has helped reimagine the data extraction workflow at TrustBridge. Earlier an entirely manual effort, bulk data extraction is now largely automated in a manner that ensures accuracy, time-savings and cost efficiency. This frees up crucial hours for TrustBridge researchers, allowing them to spend more time on analyzing extracted data.
A New Chapter in Research
The partnership between TrustBridge and Lucio represents a significant step forward in the field of regulatory research. By embracing AI-driven solutions, TrustBridge has not only improved the efficiency of their operations but also set a precedent for how technology can be leveraged to drive systemic change.