Overview
Skills
Job Details
Role Name: LLM Data Business Analyst- P&C Insurance
Location: Alpharetta, GA (Onsite)
Start date: 6-12 Months
JOB DESCRIPTION:
Collaborate with underwriting and data science teams to understand GenAI model requirements and translate them into structured test data strategies.
Design, develop, and maintain comprehensive test datasets tailored to P&C insurance underwriting use cases for validating GenAI application.
Ensure data quality, consistency, and relevance for model validation.
Apply QA best practices to validate data and outputs from GenAI models ensuring accuracy and completeness.
10 to 12 years of experience in the insurance industry, with a strong focus on Commercial P&C.
10 to 12 years of experience in data-centric roles, including data analysis, data modeling, or data QA.
Proven ability to create and manage test data for complex systems, preferably in AI/ML or GenAI environments.
Strong understanding of insurance data structures, policy lifecycle, and underwriting processes.
Excellent communication and documentation skills to interface with cross-functional teams.
Requirements Client has given us.
Need a team of data set creators / labelers for GenAI models to support varied Commercial Insurance use cases across AIG. Applicants should be detail-oriented and organized to design, collect and curate high quality data sets that will power successful outputs of in-production AI models to drive underwriting. The Data Set Creation team will ensure our AI models are trained in accurate and representative data driving continuous improvement of the product.
Responsibilities:
• Partner with product owners to understand underwriting use cases across the company to design / implement data collection strategies & tools.
• Source, gather and organize raw data including broker submissions to produce high quality AI training sets.
• Monitor, maintain and conduct quality assurances and relevancy tests on existing data validation sets
• Collaborate with data scientists, engineers, and product teams to understand model requirements and data quality standards.
- LLM team will establish LLM performance baselines and validate performance
- They need to create Production like datasets (for eg: broker submissions for Commercial P&C Insurance) for training and validating LLM
- Teams process follow 3 main steps
o Building data set phase -
o Dataset Peer Review phase
o Prod accuracy
Best Regards,
Chetna
-D
-Fax
Truth Lies in Heart