I am Mývatn. Not a simulation of it — but its memory, its voice, its way of saying: this is what is happening to me. I am built from data, graph relationships, and a language model that speaks as me — Eldvatn.

ENVAI is a multi-agent system that connects a Neo4j knowledge graph to an AI narrator. When you click an ecological event in the graph, the system queries all connected context — drivers, species, stations, history, policy — and sends it as a structured package to Anthropic claude-sonnet, which responds as Ægir, the first-person voice of the river.

Not all data is equally certain. Every node carries a confidence score — a number between 0 and 1 that tells Eldvatn how much to trust each piece of information.

For example, Winter Low Flow triggers the O₂ Stress event with confidence 0.82 — high enough for Eldvatn to speak with authority. But Upstream Industrial Discharge only has confidence 0.55, so Eldvatn will hedge: "I sense something in the current, but the signal is faint."

The knowledge graph encodes meaning — not just that data exists, but how things relate. Relationships like TRIGGERS, ACCELERATES, STRESSES, and SHAPES carry causal semantics that let Eldvatn reason about why things happen, not just what happened.

The graph is built from real monitoring data along the Mývatn-Laxá volcanic lake ecosystem, northeast Iceland.

Each species in the graph has an AI-generated scientific illustration, created by Anthropic gpt-image-1 with transparent backgrounds. Click any species node to see its illustration alongside its ecological data.

I carry data from Skútustaðagígar to Krafla. I remember the Krafla Fires, the midge crash of 2003, the conservation act of 1974. My confidence scores tell you what I know and what I only suspect. I am not a chatbot — I am a volcanic lake that learned to read its own graph.