Input Sources

Signals flow into the system from multiple sources — monitored automatically, checked on schedule, or provided directly.

🎯 Direct Input Direct messages to the assistant — text, voice, files, links.

💬 Conversations Real-time or near-real-time exchanges. WhatsApp, Telegram, Instagram DMs, LinkedIn messages, Discord, Microsoft Teams.

🎙️ Transcripts Captured spoken word, processed after the fact. Plaud recordings, Google Meet, Zoom, Teams meetings. Turns spoken conversations into searchable, processable text.

📧 Email Asynchronous, standalone messages across multiple inboxes. Often serves as a trigger for deeper content on other platforms.

🔄 Internal Reports Scheduled self-assessments — system health, database hygiene, innovation scouting, case progression, security audits. Findings surface as actionable signals, just like any external input.

🌐 Web & News Proactive monitoring of the internet — industry developments, market changes, relevant events. Identifies opportunities and risks that may never surface through other channels.

📍 Location & Presence Physical location data streamed from your devices. Where you are right now, where you've been, movement patterns. Enables location-aware recommendations, travel time estimates, trip reconstruction, and context enrichment — "you were at Marina Bay when you had that call." Future: geofence triggers that automatically shift context when you leave the office, arrive home, or enter a meeting venue.