缅北轮奸

At AM Specialty Insurance Co., Nishad Lad, head of data and software development, has been placed at the forefront of transforming the Dallas, Texas-based excess and surplus carrier’s claims processing with new technologies.

His team has implemented an array of new tools like a natural language processing (NLP)-powered bordereaux processor, and they are developing a system to automate the classification of claims based on loss descriptions. An (NLP)-powered bordereaux processor is software that uses natural language processing to automatically extract and organize data from insurance bordereaux documents.

It’s a big job, but one that Lad seems to be well prepared for.

Lad earned a bachelor’s degree in computer engineering from Mumbai University in 2018, and a master’s degree in computer science from the University of Arlington Texas in 2020. He started at AM Specialty as a software engineer. Now, as the head of data and software development, he is leading projects that are reshaping how the company interprets data and processes claims.

Claims Journal spoke with Lad about the technology they are implementing, how it’s going and what they hope to gain from their efforts. The conversation has been edited for brevity and clarity.

Claims Journal: What new technology are you integrating into your claims department/processes?

Lad: We have implemented our Automated Claims and Premium Bordereaux Processor by integrating Natural Language Processing as a part of it. Additionally, we are in the initial stages of developing a system that can understand a claim through its loss description and identify the Annual Statutory Line of Business (LOB) Code based on it in cases where it is not reported. This experimental system leverages the BART-large-MNLI model (Bi-Directional Auto-Regressive Transformers) to analyze and classify text.

Claims Journal: Why are you implementing this technology? What problem is it solving or will it solve?

Lad: Automated Bordereaux Processor solves the inconsistency in bordereaux formats and column representations, ensuring standardized data capture and interpretation. The problem arises as different companies that are involved in different stages of reporting report in different ways, terminologies which could lead to inconsistencies in the data reported

LOB Code Identification tackles the problem of manual classification of claims by understanding unstructured loss descriptions and predicting the relevant LOB code. This improves accuracy and reduces time spent on manual review.

Claims Journal: How is it working?

Lad: The NLP-powered bordereaux processor matches column content semantically to predefined data points, ensuring accurate data capture and reducing manual errors. Early tests indicated an average accuracy of 96.76%. In our beta testing, the user could process a bordereau document within a few minutes (average time of four minutes) based on the size of the data being processed. The LOB code identification system is still in a very early stage of development, and we hope it achieves a similar standard of accuracy.

Claims Journal: How do employees feel about it?

Lad: They are excited, optimistic and more accepting of the technology as it reduces repetitive tasks, enhances efficiency and acts as an assisting tool.

Claims Journal: How do you feel about it?

Lad: We are enthusiastic about both projects and their potential to transform claims processing. The high accuracy rates achieved so far are encouraging, and the experimental work on LOB code prediction represents an exciting step forward in automating complex tasks.

Claims Journal: How do will you measure the success of this technology? What are the metrics?

Lad: Accuracy rates: Maintaining or exceeding the tested average of 96.76% for the bordereaux processor. Achieving high prediction accuracy for the LOB code identification model. Processing time: Reducing the time required for manual claims and bordereaux data processing.

Error reduction: Lowering discrepancies in data interpretation and classification. Employee productivity: Evaluating improvements in employee efficiency and focus on strategic tasks.

Adoption rates: Monitoring the integration and scalability of both systems across the organization.

This story originally appeared on 缅北轮奸 sister publication Claims Journal an part of on-going series about technological and AI-related developments and advancements in claims processes and claims departments.

Garner is an intern for Claims Journal. He’s a senior at Cal State University Dominguez Hills and expects to graduate with a B.A. in journalism in 2025. He also works for the Bulletin, the student-run newspaper at CSUDH.

Topics Carriers Texas Excess Surplus