Allianz UK and University of Bristol Enhance AI Talent Development
This collaboration between Allianz UK and the University of Bristol merges academic research with practical industry applications, empowering students like Ed Davis to tackle real-world challenges through advanced data science.
Key Facts
- Allianz's partnership with Bristol enhances AI capabilities, revealing a strategic focus on innovation.
- Ed Davis's dual experience highlights talent development, crucial for retaining skilled data scientists.
- The creation of a data visualization tool indicates a competitive edge in personalized financial offerings.
- Allianz's involvement in Datathons showcases adaptability, addressing real-time industry challenges effectively.
- Collaboration with academia strengthens recruitment channels, ensuring a pipeline of skilled professionals.
Summary
Summary
Allianz UK partnered with the University of Bristol to enhance their data science capabilities through a PhD program focused on dynamic graph neural networks. The collaboration aimed to apply advanced statistical methods to real-world business challenges, resulting in the development of an internal data visualization tool that improved decision-making for customer financial product offerings.
Background
Allianz UK, one of the largest insurers in the UK, established a partnership with the University of Bristol in 2019. This collaboration was aimed at sharing knowledge and skills in data science and machine learning, initially focusing on the Faculty of Social Sciences and Law. Before the deployment, Allianz sought to deepen their understanding of complex data relationships and improve their decision support models.
Challenge
The primary challenge was to analyze how changes in customer features—such as income and default rates—could impact Allianz's decision support models. The company needed a method to quantify these changes and ensure their models remained relevant and effective.
Solution
The University of Bristol developed a suite of methods, including dynamic graph embedding, to analyze connections that change over time. Ed Davis, a PhD student, worked closely with Allianz, providing monthly updates and presenting his research to Allianz’s Innovation Forum. He helped create an internal data visualization tool that enabled Allianz to identify which customers should be offered specific financial products based on their changing performance.
Results
The collaboration resulted in the creation of an internal package that applied Ed's research findings to Allianz's data, enhancing the team's statistical knowledge and ensuring the knowledge remained within the organization. Ed published two peer-reviewed papers during his PhD, contributing to advancements in graph neural networks and related fields.
Key Insights
The collaboration demonstrated the value of integrating academic research with industry needs. Regular communication and presentations helped bridge the gap between theoretical concepts and practical applications, enhancing both the academic experience and the company's capabilities.
Customer Testimonial
“The real success for us was the creation of an internal package that applies the findings from Ed’s research in a way that it can be applied to our data in a meaningful way.” — Dr. Lorna Sumption, Lead Data Scientist, Allianz UK Personal Lines Data Science.
Entities Mentioned
Companies
Products
Technologies
People
Organizations
Key Concepts
Definitions
- Dynamic Graph Neural Networks
- Artificial intelligence models designed to analyze connections that change over time.
- COMPASS
- The EPSRC Centre for Doctoral Training in Computational Statistics and Data Science.
- graph neural networks
- A type of neural network designed to process data structured as graphs.
- conformal prediction
- A statistical technique that provides a way to quantify the uncertainty of predictions.
- network bootstrapping
- A method used to estimate the properties of a network by resampling.
Use Cases
- →Developing an internal data visualisation tool for Allianz.
- →Creating predictive models for insurance claims.
- →Forecasting pollution levels in urban areas.
- →Translating theoretical research into practical applications.
- →Engaging PhD students in real-world data science challenges.
- →Improving decision support models for customer financial products.
Frequently Asked Questions
What is the purpose of the collaboration between Allianz and the University of Bristol?
The collaboration aims to share knowledge and skills in data science and machine learning, providing PhD students with hands-on industrial experience while helping Allianz enhance its data analysis capabilities.
What benefits did Ed Davis gain from his PhD project?
Ed gained valuable experience in applying theoretical research to real-world problems, developed strong communication skills by presenting to industry professionals, and secured a job with Amazon after graduation.
How does Allianz benefit from engaging with PhD students?
Allianz benefits by gaining fresh perspectives on their challenges, enhancing their team's statistical knowledge, and establishing a recruitment channel for talented graduates.
What are Datathons and how do they contribute to the collaboration?
Datathons are intensive workshops where PhD student teams develop predictive models for specific challenges, providing students with industrial engagement experience and offering Allianz innovative solutions.
What role does the COMPASS CDT play in this collaboration?
The COMPASS CDT facilitates the training of PhD students in computational statistics and data science, fostering partnerships with industry to enhance research applicability and student employability.