IRIS gives your AI agent a real eye. Camera, screen-share, file drops, screenshots — every visual input becomes live context the agent can reason on, act from, and remember.
Point your phone at a whiteboard, a menu, a circuit board, a contract — IRIS reads it in real time. Streaming inference at sub-second latency.
Use the iOS share-sheet to send any screenshot into a live IRIS session. The agent sees what you saw and resumes where you stopped.
Hand over a 10-second screen-record and IRIS frame-samples it to follow the action. "What just happened in this app?" answered in under a second.
PDF, slides, diagrams, photos. Drop them into the conversation. IRIS extracts, understands, and acts — no separate upload flow.
Everything IRIS sees gets indexed in your private memory. "Find the receipt from last month with the rainbow logo" — done.
Vision + text + voice in the same turn. Ask about an image, follow up by voice, get a code snippet back. One agent, one context.