Inspiration
When one of my best friends passed, all the memories I had of him took on a different, bittersweet tone. It reminded me of how important photos can be in preserving those shared moments, and it sent me on a wild goose chase to gather all these traces I knew were in my gallery. It took a long time—
We think it shouldn't.
Photos are important. And yet, the most important photos get lost in the thousands of random snaps that drench your camera roll. With Echo, we wanted to develop a way to easily bring these invaluable memories back to the surface, and collect (curate) it for our closest people to see & remember.
What it does
Echo is a personal AI memory curator that combines a UI with memories front-and-center with a robust search-and-share engine backed by the raw power of LLMs. Core features:
- Comprehensive filtering and search options with deep contextual understanding
- Seamless cloud integration with major cloud storage platforms (Google Photos, iCloud, etc.)
- Intuitive & smart album curation, collaboration, and sharing process
- Community discovery with location or event albums
As an inherently human & connection-driven tool, Echo invites users to refine & redefine their own history into collections that bring us together & helps us hold on the connections that have made us grow.
How we built it
Our multimodal photo analysis & semantic search engine leverages the Google Gemini 2.5 Pro model to generate rich metadata on photo upload and deliver powerful semantic understanding on search queries.
Challenges we ran into
Our most significant challenge was figuring out how to manage & store photos efficiently proved challenging, particularly with respect to optimizing LLM token usage and computational costs.
Accomplishments that we're proud of
We are proud of designing an attractive & intuitive UI that has an inherently wholesome and human-centered identity. Artificial intelligence is used to promote (and not replace) the connected human experience, bringing together experiences & feelings that encourages healthy reflection & gratitude.
What we learned
We learned that design & ideation is sometimes as difficult as implementation itself. In addition, knowing when to stop, and when to prioritize other elements (core features, for example). Building, especially in a short time period, must be purposeful!
What's next for Echo
Our hopeful next steps include onboarding real users (!), optimizing flow, and developing more options (e.g. video support) for people to share, discover, and connect through shared experiences. Potentially scaling with one of our local incubator tracks.
Log in or sign up for Devpost to join the conversation.