Zoo Labs · applied research
AI for conservation
Machine learning applied to protecting endangered species — monitoring, prediction and the conservation strategies they support.
Open weights
The ZenLM framework
ZenLM is trained on conservation data, species behaviour and environmental factors, and published with its weights so the results can be checked.
- Species identification from camera-trap images with 98% accuracy
- Predictive modelling for migration patterns and habitat change
- Real-time poaching threat detection and alerting
- Ecosystem health monitoring through satellite imagery analysis
Impact metrics
- Wildlife images analysed
- 15M+Wildlife images analysed
- Species identified and tracked
- 2,300+Species identified and tracked
- Reduction in response time to threats
- 87%Reduction in response time to threats
What the lab works on
Research areas
Computer vision
Image recognition for species identification, behaviour analysis and population counting from drone and camera-trap footage.
Predictive analytics
Models that predict habitat loss, climate impact and migration change, to enable proactive conservation.
Natural language processing
Analysis of research papers, field reports and social media to track conservation trends and identify emerging threats.
Acoustic monitoring
Audio analysis that identifies species by their calls, monitors ecosystem health and detects illegal activity.
Genomic analysis
Machine learning applied to sequencing data to understand genetic diversity and guide breeding programmes.
Ledger integration
Verified wildlife data written to a public ledger, for transparent and tamper-evident conservation records.
For developers
Conservation AI API
The conservation models are reachable at api.zoo.network.
Endpoints
- POST/v1/species/identify
- POST/v1/habitat/analyze
- GET/v1/threats/monitor
- POST/v1/migration/predict
- POST/v1/audio/classify
Collaborate
Join the AI conservation work
Partner with us to develop and deploy models that make a measurable difference in wildlife conservation.