High-volume WhatsApp support is where most ERPNext teams quietly lose the thread. Messages land on one shared phone or a single BSP number, several agents answer from the same screen, and there is no record of who replied, what was promised, or how long the customer waited. ERPNext ships CRM, Helpdesk and a generic Notification/Communication layer, but it has no concept of a live multi-agent WhatsApp queue: no per-conversation ownership, no first-response SLA clock, no quick-reply library, and no way to see the whole inbox as a workload you can distribute. Teams end up bolting on a third-party inbox that lives outside ERPNext, so the conversation history never reaches the Customer, Lead or Issue it belongs to.

We build a whatsapp_team_inbox Frappe app that puts the entire shared inbox inside ERPNext itself. Inbound messages arriving via your WhatsApp Business API provider (Meta Cloud API or a BSP such as Twilio/360dialog) hit a whitelisted webhook endpoint (@frappe.whitelist(allow_guest=True)) that verifies the provider signature and writes each message into a WhatsApp Conversation DocType with child WhatsApp Message rows. Conversations carry owner, status, tags, linked party (Customer/Lead/Contact via Dynamic Link) and SLA fields. A custom Desk workspace and a real-time list — refreshed over Frappe's frappe.realtime/socket.io channel — give agents a true shared inbox instead of a static report.

Assignment and routing are driven by ERPNext's own primitives. hooks.py doc events (after_insert, on_update) run server-side logic that auto-assigns new conversations by round-robin or by team/language rules, applies ToDo/_assign ownership so it shows in each agent's assignment panel, and stamps the SLA due timestamp. Canned replies are a Canned Response DocType with per-team visibility and variable substitution (e.g. {{ customer_name }}), surfaced through a client script slash-command in the reply box. A scheduler event (registered under scheduler_events in hooks.py) sweeps open conversations every few minutes, flags first-response and resolution SLA breaches, and escalates or notifies team leads. Everything respects ERPNext permissions: we ship dedicated roles and role profiles, so agents see only their team's queue while supervisors get the full board and the metrics. All conversation and message data is reachable through the standard Frappe REST API and additional whitelisted methods, so your reporting or a future dashboard can read it cleanly.

Because this is build-to-order, we tailor the routing rules, canned-reply library, SLA policy and party-linking to how your team actually works rather than shipping a rigid generic add-on. After a short scoping call we confirm the exact DocTypes, roles, provider integration and SLA definitions in writing, then build on a staging bench. Typical delivery is one working week from confirmed scope, depending on how many teams, SLA tiers and provider quirks are in play. We install and configure it on your ERPNext v15 or v16 site, run UAT on staging with a rollback plan before touching production, hand over the full git repository and documentation, train your agents and supervisors, and stay on for a post-go-live support window. There is no instant download — this is a system we build, install and stand behind.