DeskcommCRM setup and AI prospecting for WhatsApp

This Portuguese-language tutorial introduces DeskcommCRM, an open source sales CRM for WhatsApp that the presenter runs on a VPS. He demonstrates company searches and automated outreach, alongside an inbox, sales funnels and configurable agents with knowledge bases.

The partial setup demonstration uses SSH access, a domain, Supabase and an AI provider API key, with the remaining steps provided in the author’s linked material. The presenter recommends 4 GB of RAM when using Supabase Cloud, or 8 GB when hosting Supabase on the same server. Supabase stores leads, conversations and knowledge data. This self-hosted deployment still uses external services in the demonstrated configuration; the presenter selects OpenAI during installation. At the time of recording, he says Portuguese and Spanish are available, with English and Chinese planned.

The prospecting example searches for aesthetic clinics in Divinópolis, sets a company limit and caps search spending. Apify supplies the scraping API and requires a personal token. Search results and service use vary with the configuration.

For outreach, he selects an existing AI agent, an outgoing WhatsApp number and a sales funnel. He configures a daily contact limit and a ten-minute interval to reduce blocking risk. The interface warns that finding contacts does not establish consent. The demonstration ends with an initial message, an opt-out instruction and campaign tracking for delivery and replies.