Inspiration

Birds are important indicators of environmental health, yet their populations face growing threats from climate change, habitat loss, and human activity. However, bird sighting data is scattered across multiple sources, making broader trends difficult to explore. We wanted to bring this information together to make wildlife data more accessible!

What it does

BirdBrained is an interactive wildlife conservation platform that brings bird sightings and climate data together in one place.

With BirdBrained, users can:

  • Explore breeding-season sighting trends across multiple bird species and regions.
  • Visualize geographic and temporal shifts through an interactive 3D globe.
  • Identify unusual changes in sighting patterns, with flags where trends lack statistical clarity.
  • Chat with an AI analyst to generate evidence-backed hypotheses and uncover potential insights.
  • Upload and visualize their own custom datasets.

Whether you're a researcher or simply curious about birds, BirdBrained makes wildlife data easier to explore.

How we built it

We built the frontend with React, TypeScript, Globe.gl for interactive 3D visualizations, and Lucide for icons, paired with a Python and FastAPI backend. We integrated the GBIF API (including eBird and iNaturalist observations), USGS Breeding Bird Survey data, and NOAA climate records to bring together wildlife and environmental information.

We also integrated Gemini API to power our AI analyst, which helps users interpret trends and investigate potential patterns. We used ChatGPT and Claude Code as programming assistants throughout development.

Challenges we ran into

Bird sightings aren't direct population counts, and citizen-science records have grown dramatically over time. To account for fluctuating observer activity, we normalized individual species counts against total recorded sightings, so users see relative trends rather than artificial spikes.

We also had to consolidate datasets from multiple sources with different formats and geographic coverage.

Accomplishments that we're proud of

We're proud to have combined fragmented wildlife and climate datasets, interactive 3D visualization, and an evidence-backed AI analyst into one platform that serves both researchers and the general public.

What we learned

We learned to distinguish observational data from actual population measurements and the importance of interpreting trends within the limitations of the available data. We also explored how AI can support scientific research by helping analyze data and identify patterns without replacing researchers' expertise.

Built With

Share this project:

Updates

Submission history