Deskcomm CRM: Open-Source AI Sales OS With WhatsApp Agents

TL;DR

TL;DR: Deskcomm CRM (DeskcommCRM) is a self-hosted AI sales OS. It bundles a CRM with native AI agents and a WhatsApp integration via WAHA. The project positions itself as an open source alternative to Kommo, Octadesk, and other commercial sales CRMs that lock customer conversations inside their walls.

What Is Deskcomm CRM?

The repository description: “Open-source AI sales OS — self-hosted CRM with native AI agents + WhatsApp (WAHA). Open alternative to Kommo, Octadesk & [others].” That positions the project clearly. It is not a clone of any one product — it is a stack: a CRM, AI agents that live inside the CRM, and a WhatsApp channel that the agents can drive through WAHA.

The combination matters. Most open source CRMs stop at contacts and deals. Most open source chat bots stop at one channel. Deskcomm CRM tries to glue the full sales loop together — conversations on WhatsApp, structured deal state in the CRM, and AI agents that can move deals forward on either side.

How It Works

The TypeScript codebase targets a self-hosted deployment. The architecture implied by the description and topics — ai, ai-agents, chatbot, crm — is roughly:

• A CRM core with contacts, deals, pipelines, and the usual sales metadata.
• An AI agent layer that can read CRM state and act on it (draft messages, summarize threads, suggest next steps).
• A WhatsApp connector via WAHA, the open source WhatsApp HTTP API bridge. That means the AI agents can read incoming messages and reply on the WhatsApp channel without leaving the stack.

Self-hosted means you control the data. For sales teams that handle sensitive conversations, that’s the actual differentiator versus hosted CRMs.

Repo Facts

• Repository: melgarafael/DeskcommCRM
• Language: TypeScript
• License: MIT
• Stars at time of writing: 1,371
• Topics: ai, ai-agents, chatbot, crm

When To Use It

Deskcomm CRM is the right tool when:

• You run a sales team that lives on WhatsApp and you want to keep that data in-house.
• You need an AI agent layer that can act on CRM state, not just summarize documents.
• You are willing to self-host a TypeScript stack and operate a WAHA bridge.

It is not a hosted SaaS. There is no company behind it selling you a managed plan, and there is no SLA. If you need a vendor to call when the AI agent breaks, this is not the right stack yet.

Source and Accuracy Notes

• Project page: https://github.com/melgarafael/DeskcommCRM
• Source repository: melgarafael/DeskcommCRM (verified via GitHub API html_url)
• License: MIT (verified via GitHub API license.spdx_id)
• Stars at time of writing: 1,371
• WAHA, Kommo, and Octadesk are referenced as third-party products; verify their current licensing and operational status before relying on the integration in production.