Searchem

Picture of Searchem AI

Inspiration This idea came to me while I was driving. I wanted to throw on some music that matched how I was feeling — but I couldn’t really do that safely while on the road. I also realized I didn’t have a go-to playlist that just got the vibe, whether I was feeling good, alone, heartbroken, with friends, or just cruising. So I thought, what if an app could read your mood and play music that fits? That’s how Searchem started.

What It Does Searchem listens to how you’re feeling — literally. You can type or say your mood, and the app figures out your emotion and pulls up a Spotify playlist to match. It makes picking music easy, especially when you're not sure what you want to hear but know how you feel. I also listed some common emotions to help you out if user is stuck.

How We Built It I built this with React and added voice input using the browser’s Speech Recognition API. For figuring out emotions, we used a sentiment analysis tool and combined that with a list of emotion-related keywords. Based on what the app picks up, it shows a fitting Spotify playlist using their embed feature.

Challenges We Ran Into At first, it was tough to match what people were saying to a specific emotion. Sometimes moods are complicated or don’t fit one word. I had to tweak the logic a lot and test different ways to recognize how someone was actually feeling. Eventually, combining sentiment scores with keyword detection worked pretty well.

Accomplishments That We're Proud Of I am proud that the app works — it actually listens, understands your mood, and plays music that fits. Getting the emotion detection to feel accurate and making the experience feel personal was a big win for us.

What We Learned I learned how powerful sentiment analysis can be with natural language, how to combine tools like voice input and Spotify together, and how important little UX details are like showing emotion suggestions to guide the user.

What's Next for Searchem Next, I'd love to make it smarter maybe use machine learning to understand moods more deeply, or let users create and save their own custom playlists based on how they feel. There’s a lot of potential for personalization, and we’re just getting started.