Data Scientist

Jobtailor; Alabama; On-site; USD 85,000 - 120,000; Full time

Job summary

Kemper is seeking a data scientist to build predictive models and applied LLMs for fraud detection, subrogation, vehicle damage severity, and claims quality management. The role emphasizes automation of modeling processes and improving data infrastructure.

Strong Python skills and a quantitative graduate degree are required, with 2+ years in analytics preferred. Experience with AWS, NLP, and LLMs via API is considered a plus.

Qualifications

• Proficient in Python with libraries like scikit-learn, pandas, numpy, scipy.
• Experience with predictive modeling techniques including GLMs, trees, ensembles, regularized models, clustering, and neural nets.
• Some experience with AWS is a plus.
• Some NLP experience is preferred.
• Some experience with LLMs via API is preferred but not required.
• Ability to work with various data formats (RDBMS, delimited text, data frames, JSON).
• Excellent communication skills, able to translate results for non-technical audiences.
• Graduate degree required in a quantitative field (e.g., Math, Stats, CS, Physics, Economics, Electrical Eng).
• 2+ years’ experience in data science/predictive analytics preferred.
• Self-directed with minimal supervision.

Responsibilities

• Work as part of a team to build predictive models and applied LLMs to support claims operations (fraud detection, subrogation, damage severity, adjuster triage, claims quality).
• Develop and automate predictive modeling processes deployed across the organization to solve analytics needs.
• Contribute to improving modeling capabilities and data infrastructure by tweaking existing processes.
• Monitor deployed solutions and address issues as needed.

Skills

Python programming

Communication skills

Self-direction

Education

Graduate degree in a quantitative field

Tools

scikit-learn

pandas

numpy

scipy

AWS

LLMs via API

NLP

Job description

Job Responsibilities

• Work as part of a team to build predictive models and applied LLMs designed to support the Kemper claims operation in areas such as fraud detection, subrogation identification, vehicle damage severity, adjuster triage and claims quality management.
• Develop and automate predictive modeling processes that can be deployed through the organization to solve reoccurring analytics needs.
• Contribute to the improvement of internal modeling capabilities and data infrastructure by suggesting and applying tweaks to existing processes.
• Monitor deployed solutions and escape as needed.

Requirements

• Proficient in Python programming language (scikit-learn, pandas, numpy, scipy, etc.).
• Proficiency and experience with several modeling techniques such as generalized linear models, decision trees, ensemble learning, regularized models (ridge/lasso/nets), clustering, and neural networks.
• Prior experience working with AWS is a plus.
• Some experience with natural language processing is preferred.
• Some experience working with LLMs via API is preferred but not required.
• Ability to work with various data formats. This includes relational databases, delimited text files, data frames, and JSONs.
• Excellent overall communication skills, especially the ability to translate technical results for non-technical audiences.
• Graduate degree required in a quantitative field: Mathematics, Statistics, CS, Physics, MIS, Economics, Electrical Engineering, etc.
• 2+ years’ experience in a data science/ predictive analytics environment preferred.
• Self-directed and able to work with little supervision.

Core Competencies

Demonstrates expertise in Python programming and predictive modeling techniques to develop and automate analytics solutions. Capable of translating complex technical results for diverse audiences while contributing to data infrastructure improvements.