Inspiration

We don’t protect what we can’t see. ReefPulse makes ocean change visible and interactive so curiosity leads to action.

What it does

ReefPulse is an AI-powered ocean intelligence app that combines safety insights, marine-life discovery, anomaly alerts, and community reporting in one place. It turns complex ocean data into clear guidance for safer, smarter coastal decisions.

How we built it

We built a React + FastAPI platform with ML services deployed on AWS SageMaker. The backend routes forecasting, anomaly, and species inference, while the frontend delivers maps, dashboards, auth, and community UX.

Challenges we ran into

We hit endpoint deployment issues (health checks/timeouts), cross-account IAM complexity, schema alignment across services, and frontend reliability issues with external assets.

Accomplishments that we're proud of

We delivered a working full-stack product on AWS, deployed and integrated multiple ML endpoints, and built a reliable backend architecture with clear handoff docs for multi-account collaboration.

What we learned

Production ML success depends on serving contracts, observability, and secure credential workflows as much as model quality.

What’s next for ReefPulse

We’ll harden model serving latency/reliability, complete fully live multi-model orchestration, expand real-time ocean data, and pilot ReefPulse with coastal communities.

Share this project:

Updates