Our project worked with The Hartford and focused on researching and prototyping Differential Privacy (DP) techniques for use in insurance data analytics. We began by conducting a comprehensive investigation into the fundamentals of DP, including its mathematical foundations, privacy-utility trade-offs, and its relevance to the insurance industry. This included reviewing and comparing existing tools, datasets, and use cases, as well as exploring the broader ethical and regulatory implications of privacy-preserving data analysis. To bring our research into practice, we developed a full-stack prototype system showcasing DP in action. We utilized SUMO (Simulation of Urban Mobility) to generate synthetic vehicle telematics data, representative of real-world insurance applications. This data was then processed using OpenDP’s PyDP library to apply differential privacy to sensitive data points, introducing calibrated noise while preserving overall utility. We computed vehicle risk scores on both raw and differentially private datasets, enabling a comparative analysis of the privacy-utility trade-off. These results were visualized using heatmaps and other interactive graphics. To bring our work all together, we created an interactive website to host the prototype and make Differential Privacy accessible to a broader audience. The website allows users to explore our dataset, simulate custom vehicle trips, view differentially private outputs, and learn about DP’s relevance and benefits in real-world insurance contexts. The site also features educational resources explaining DP concepts and its critical role in enabling secure, ethical analytics in regulated industries.