Lisa AI – Mobile app for matching candidates and partners in different areas

The client got in touch with us and presented a ready product to us that required a full-fledged iOS app. On the basis of the KMM module and the Figma design that the client provided to us, we developed a native solution using KMM, SwiftUI, UIKit, and Lottie animations.

The main problem was that before we got involved, the product had been written in UIKit and we had to adapt the app for SwiftUI. We decided to go with that framework due to its layout speed. We needed to make the app the same as it was when it was designed, and we matched it pixel for pixel.

Like on Tinder, profiles listed users’ answers to questions regarding what they did and didn’t like. One of the most troublesome aspects of development was implementing that profile stack, since SwiftUI had a tough time handling the animations. The profiles had a lot of animations and performance would emerge as the main issue. Everything worked smoothly on the new devices, but as soon as the app was launched on an old device, their battery would overheat, glitches would happen, and animations would get jerky and sluggish. The project gained a second wind after the custom stack was written.

We implemented parsing of a long animation for certain pieces (markers) and logically connecting those pieces into a complete picture. At the same time there is interaction with the user: when they press a certain button, the animation smoothly switches, displaying the next marker.

Profiles featuring questions regarding user ratings are configurable as from Junior- to Senior+. Each marker features its own animation. Junior- utilizes the visual icon of a baby holding a bottle while a Middle Developer is represented by a young man wearing glasses, sitting at a computer. There are also profiles featuring regular buttons to visit another user’s account or set up filters as well as profiles featuring user feedback recording. The user would press on an animated button containing a microphone, the pulse would start to vibrate depending on the volume of the conversation, and a sound animation would launch (like on old digital stereos, which went along with the beat). The client planned to use these profiles to collect feedback from users. 

Some profiles the users add themselves. For instance, they can upload several togglable photographs. On top of that, many buttons in the app are tactile – when they’re touched the device’s vibration motor is activated, giving a feeling of smoothness. But the most interesting function is registration, built on the idea of communication between the user and the app. When the person logs into the app and completes onboarding, Lisa herself, the company’s mascot, communicates with him. By the way, the mascot is always on the main page and from time to time it plays various Lottie markers (animations sent from the backend). There are always new things she can do.

Thanks to the broad range of functions, performance, complex matching algorithm, and LISA artificial intelligence technologies, the app assembles the perfect couples sharing common interests.