See the night.
Keep the decision human.
Three ways to explore one evidence record.
Begin with the connected story, read the shift-level picture, then ask a focused question. Every path leads back to source footage.
- 01Node graphFollow every connection
Trace animals, cameras, observations, clips, events, and keeper outcomes through the connected evidence graph.
Explore the graph - 02AnalysisRead the whole shift
Compare behavior, rule events, coverage, and review status in a plain-language operational summary.
Open analysis - 03AI chatbotAsk the evidence
Ask what an animal was doing and jump back to the cited moment without losing the selected camera context.
Ask a question
Run one minute of night watch.
Pick a moment, scan the evidence, then decide how it should be routed. Follow each observation through baseline context, the first matching rule, and a human decision.
Watchful and human-led.
ZooVision observes, preserves evidence, and routes factual checks while keeper judgment remains final.
Timestamped behavior with the source clip attached.
Explicit Python rules assign severity and preserve rule_fired.
Every routed check ends with a clear human decision.