Everything ChatterOS does.

Designed by someone who actually works the shifts. Every feature solves a real headache.

OCR auto-capture

Reads Infloww notifications off-screen using on-device OCR. No browser extensions, no platform integrations to break. Fires within 500ms of a sale appearing.

🧠

Smart routing

Maps detected sales to the correct model using your team's configured roster, fuzzy match, customer-handle memory, and team-suffix fallback. Unrouted sales go to a quarantine bucket for one-click assignment.

Real shift presence

Start Watching = clocked in. Finish Shift = clocked out. Heartbeats every 30 seconds. Stale clock-ins auto-expire after 2 minutes of silence.

⚖️

Equal shift splits

When 3 chatters share Model X in Shift 2, every sale auto-splits 3 ways. Splits lock at sale moment — late clock-ins only share future sales, never retroactive.

📊

Live dashboard

Owners see every team and every chatter in one grid. Net, subs, ratio, online status — all auto-refreshes every 5 seconds. Filter by shift in one click.

📅

Day-off requests

Chatters request time off from the desktop app. Owners approve or deny from the dashboard. Quotas track per chatter, per month.

✉️

Two-way messaging

Chatter sends a request → owner sees a red badge → approves with a note → chatter sees the decision next time the app refreshes. No DMs lost.

💰

Per-chatter payout tracker

Click any chatter to see their full month: total net, your-payout, per-day breakdown, team vs cover. Replaces every Excel sheet.

🛡

Multi-tenant security

Postgres row-level security ensures one agency can never see another's data. Per-chatter activation keys. Email + password owner login.

☁️

Offline-first sync

Tracker keeps working without internet. Sales batch locally and flush when connectivity returns. Manual "Push now" button for impatient moments.

One-click resync

If local and cloud ever drift, "Reset cloud" wipes today's data and re-pushes from the local tracker. Source of truth always restorable.

🔍

OCR snapshots

Every detected sale saves the captured image to a local folder. Misroute? Open the snapshot, see exactly what the OCR saw, train the matcher.

Get a demo