Lead Data Scientist

Summary

Guidewire-Cyence is searching for a Lead Data Scientist to join us on our mission to transform cyber-insurance with the industry's leading cyber risk platform. Cyber is the #1 threat to US national security and the fastest growing line of business in P&C insurance. Our work here makes a meaningful impact in this space! You will report to the Director of Risk Modeling and you will also be involved in several other initiatives across the broader Data & Analytics Division.

Who We Are, What We Believe, & What We Build
Guidewire is the AWS of insurance. As the market leader, 540 insurance companies run on our mission-critical platform. Every second, we support underwriters crafting policies and agents settling claims. We believe that making a great decision should not require 100 in-house data scientists. Our products range from cyber risk quantification to potent ML sandboxes. We are a post-IPO company with the vision to redefine insurance.

Who You Are
You are a detail-oriented, passionate data scientist who builds models that hold up in production. You have strong experience developing GLM-based and stochastic frequency-severity models similar to those used in natural-catastrophe portfolios (e.g., severity tails, dependency structures, event simulation, and portfolio aggregation). You are familiar across AWS tools and environments — SageMaker for training/hosting, EMR/Spark for large-scale feature pipelines, Glue for orchestration, Lambda and Step Functions for serverless workflows, and Athena/Redshift/S3 for analytics and storage. You write clean, maintainable Python/SQL and Spark code with unit/integration tests, versioning, and CI/CD, and you partner closely with engineering to deploy models as services or batch pipelines. You balance statistical rigor with pragmatic delivery, communicating uncertainty and model trade-offs to both technical and executive audiences. Experience in CAT modeling is a plus!

Job Description

Responsibilities

• Develop, calibrate, validate, and stress test probabilistic cyber risk models for the (re)insurance and financial services markets.

• Develop and implement methodologies to quantify the financial impact of cyber risk on single entities as well as large insured portfolios, with a specific focus on Targeted attacks, cloud outages, and digital/physical supply chain-related events.

• Explore different data sources to come up with modeling features and assumptions to enhance our set of probabilistic risk models.

• Define and implement a globally consistent best practice process for data validation, feature selection, and modeling of catastrophe exposures.

• Integrating/automating modeling processes in our production pipeline to feed our platform.

• Communicate results to internal and external stakeholders. Engage and collaborate directly with clients on a deep technical level.

• Collaborate with other leaders across the Analytics organization, including Product Management, Machine Learning and Data Engineering Teams, UX/UI Engineers, and Client Engagement Leaders.

• At Guidewire, we foster a culture of curiosity, innovation, and responsible use of AI—empowering our teams to continuously leverage emerging technologies and data-driven insights to enhance productivity and outcomes.

Qualifications & Requirements

• Education: PhD or MS in Applied Mathematics, Statistics, Computer/Data Science, Engineering, or a related quantitative field.

• Experience Level: 7+ years in applied statistics/data science or quantitative risk modeling, with demonstrable delivery of scalable models that run in production.

• Modeling Expertise: Proven track record building and validating advanced predictive GLM and stochastic frequency–severity models (e.g., Poisson/NB, Tweedie, GLMMs), tail modeling/EVT, dependency structures (e.g., Gaussian copulas), Monte Carlo simulation, and portfolio aggregation—ideally in insurance or catastrophe-risk contexts. Familiarity with advanced sampling methodologies for multi-variate distributions and the usage of  common libraries such as MLFlow is a plus.

• Methodological Rigor: Strong command of feature engineering, uncertainty quantification, calibration/back-testing, A/B or hypothesis testing, and explain ability for executive and regulatory audiences.

• Data & Engineering: Proven ability to work with terabyte-scale datasets using distributed processing frameworks and cloud data platforms. Proficiency in SQL and Python for data manipulation, ETL, and large-scale analytics. Experience with modern data warehousing solutions (e.g., Iceberg, Redshift, or their equivalents).

• Cloud & MLOps: Profici