Faind
AI Networking
Background
Faind is a mobile app that uses AI to automatically connect you with people who share similar interests and aspirations. The project was at early stage of its development as part of the Antler incubator when I joined to help design its core features and bring the app to its first release.
Role
UX/UI Designer
Platform
iOS / Android
Year
2023
Industry
Networking
Problem
Currently networking apps require users to scrolls to hundreds of unfiltered matches thus making you spend a lot of time for no potential results. In Faind, we thought to use AI to automate this process. Users can tell the AI all their preferences and let it work in background for you.
Test idea
before design
Before committing time and building a fully separate app we needed to test the hypothesis that proposed AI solution will work in practice. To do that developers created a chatbot in Telegram that would work on the same idea and that we could use to test the matching technology with real people.
User Flow
After we tested the hypothesis, I started to work on a user flow for the app. The key aspect is to minimize the amount of actions user has to make to start getting matches.
Key Features
In my designs, I focused primarily on prompt creation and connections screens. Users can start new prompts by either going through preset options or use chat window to communicate with a bot directly. This is to make sure users can communicate with AI in a comfortable for them way.
UI
Considerations
A significant amount of time was dedicated to designing prompt presets and app notifications. The goal is to reduce the effort required to create a prompt, and since the AI operates in the background, notifications are essential to ensure users don't miss potential connections.
UI Kit
After the core design was complete, I created a UI Kit with all the styles and component that developers replicated in their dev environment for further improvements to the platform.
Results & Lessons Learned
After releasing the MVP, I gathered feedback on user behavior. Users tend to prefer creating custom prompts, but often feel overwhelmed when writing their own text. Additionally, there is a trust issue regarding sharing personal information with a new startup and AI.
Lesson 1: Test the tech before design
We saved a lot of time testing what we could do in Telegram before designing separating app with all the features.
Lesson 2: "In order to move forward, you have to leave something behind"
Although I had a clear vision for the final design that I wanted to realize from the start, working on an early-stage startup taught me to let go and prioritize tasks at hand.
Lesson 3: Not everything sticks
While working in team on this project was great experience, the product itself couldn't find its niche in a market. After some time it was decided to shut down the app and take on other pursuits.