The true beauty
of AI.
dew ai
Beauty & Cosmetics
B2C
Dew is an AI platform that helps you keep your beauty products organized, learns about your preferences, and provides you with suggestions regarding future products according to your personal taste and skin type.Lexogrine has been part of the journey from the simple idea to the whole application development.
Making sure to remember all the beauty products you have can be a struggle so it was a time to make it easier thanks to AI.
Challenges
We aimed to revolutionize beauty product organization and recommendations, navigating the complexities of AI development to meet unique industry challenges.
Finding the right email
Finding emails including receipts from different shops to build the database of products.
Designing the model
Creating the right AI model to efficiently search for the right emails and cataloging the products.
Creating appealing UI
Designing a beautiful user interface and UX to match the beauty market and user expectations.
Creating the right suggestions
Making use of the users’ product databases to generate beauty product suggestions tailored to each user’s specific needs.
Solution
Our solution harmoniously integrated AI's analytical power with a user-centric approach.
Development
Model trained on hundreds of emails
We trained the model on hundreds of different emails with a vast range of order confirmations to allow it to create a pattern for searching for the right messages.
Right AI model + solid codebase
We had to make sure we were building both an efficient AI model and a scalable codebase for the whole application. We used TypeScript as the core language, React Native for the mobile application, and Node.js for the backend. Infrastructure and computing power has been provided by AWS.
Designing for user delight
The beauty market has its own specifics that we had to follow to create a familiar environment for the users. We had to base the design on delicate colors and icons with a specific character to give the application a specific style.
AI-based recommendations
A dedicated AI model was created to analyze each user’s product catalog and predict what other beauty products and cosmetics can be useful - and even search for discounts!
eCommerce patterns usage
We incorporated proven eCommerce patterns into Dew AI’s design, ensuring a smooth, intuitive shopping experience enhanced by AI-driven personalization.
Tech Stack
Crafting Dew AI, we wove together TypeScript's agility, React Native's versatility, and AWS's robustness, creating a tech tapestry fit for the beauty industry’s digital transformation.
TypeScript
React
React Native
MySQL
Node.js
PyTorch
AWS
Minimum Lovable Product
One of the biggest challenges we needed to face while designing and implementing the app, was the fact that the app was in the MVP stage. By choosing the right feature scope and design choices, we were able to create an app that was useful and enjoyable.
Backend technology, prepared for heavy lifting
The app's backend, built with TypeScript and NestJS, enabled seamless switching between frontend and backend development. The usage of PyTorch proved useful in preparing a solid pattern recognition module for detecting and cataloging users’ beauty products.
Results
The team knew from the very beginning that we needed to ship fast as we were building an MVP to find product-market fit. Thanks to the usage of the elements with which we have a lot of experience - like TypeScript, React Native, Node.js, Next.js, and AWS - Lexogrine could allocate most of the time for AI development to design both efficient and scalable solutions.
Within just 2 months the initial version was released for closed beta testing gathering very positive reactions.
