Ulta Beauty App AI/AR Skin Analysis

In 2019 I was working directly at Ulta Beauty on the AI/AR Skin Analysis feature in the Ulta Beauty app. Users allowed camera access and answered a few questions, and the feature returned a personalised skin report diagnosing conditions like redness, breakouts, dark spots, and fine lines, with curated product recommendations and filtering options.

My role was building the backend middleware layer in Node.js and TypeScript: a service that sat between the mobile app and the external APIs, integrating proprietary computer vision technology for skin analysis and a product recommendation API to serve personalised results. The service was deployed on Google Cloud for scalability and reliability.

The Skin Analysis feature became a flagship part of Ulta Beauty's digital innovation strategy, covered by Forbes, Allure, Google Cloud Blog, Chain Store Age, and Cosmetics Business. The feature saw strong adoption through the app and was highlighted as a key example of AI-driven personalisation in retail beauty, with Ulta continuing to develop and expand it.

Node.js, TypeScript, Webpack, Google Cloud. Backend middleware architecture integrating computer vision and product recommendation APIs.

Backend Development: Sole developer on the Node.js and TypeScript middleware layer, building the full service infrastructure from the ground up.

API Integration: Integrated proprietary computer vision technology for skin analysis and an external product recommendation API to deliver personalised results per user.

Deployment: Deployed and maintained the service on Google Cloud, optimising for performance and scalability.