Stickered Situation

Inspiration

A crying cat sticker can mean “that’s hilarious” to one person and look like a cry for help to another. We wanted to help older adults feel included in conversations where cartoon animals somehow do most of the talking.

What it does

Stickered Situation uses sticker images and recent conversation context to generate short explanations of their likely meaning. It’s designed to explain unfamiliar humor and reactions privately, without making someone ask the entire chat to explain the joke.

How we built it

We built a Python backend with SQLite for conversation history, OpenAI’s multimodal API for interpretation, and a Photon API as a messaging adapter. The prototype combines an image with recent messages and user-recipient data to produce an explanation.

Challenges we ran into

Native sticker media wasn’t available in our messaging tests, so we added support for stickers sent as PNG attachments. We also had to handle sarcasm and conflicting interpretations without presenting guesses as facts.

Accomplishments that we're proud of

We’re proud of turning a ridiculous premise into an integrated, locally tested prototype. We built conversation-aware interpretation, private delivery routing, and safeguards against processing the same message repeatedly.

What we learned

Recognizing a screaming hamster and understanding why someone sent it are very different problems. We learned how much context matters and why a helpful explanation sometimes needs to say “probably.”

What's next for Stickered Situation

Next, we want to verify the complete live messaging flow and test explanations with older adults using conversations they consent to share. We also want to evaluate multilingual explanations and expand native-sticker support as platform access allows.