n8n Cold Email Automation: OpenAI + LinkedIn + Qdrant

Automate your multi-step cold email campaigns—personalized with LinkedIn, powered by OpenAI, and grounded in a Qdrant knowledge base

This n8n Cold Email Automation: OpenAI + LinkedIn + Qdrant orchestrator workflow automatically schedules and sends the 1st through 4th touches, enriches lead context from LinkedIn via RapidAPI, generates hyper-personalized outreach with OpenAI, retrieves accurate product details from Qdrant, and tracks follow-ups in NocoDB.

What this workflow does

• Schedules cold email sequences using Schedule Triggers, launching the 1st, 2nd, 3rd, and 4th emails at staggered times during the week.
• Pulls campaign context (lead lists and sender/company information) from NocoDB.
• Runs a sub-workflow to check inbox sending limits (daily/weekly) before sending the next email.
• For first-touch emails, fetches the prospect’s LinkedIn profile and recent posts via RapidAPI (professional-network-data endpoints).
• Uses LangChain Agents with OpenAI to analyze LinkedIn data, create a research briefing, and draft a highly personalized email.
• Retrieves product details from Qdrant (vector store) to keep generated content accurate and consistent.
• Sends via a “Send Emails” sub-workflow, then writes timestamps and sent-message records back to NocoDB to power follow-up tracking.
• For later sequences, generates context-aware follow-ups using past email history and company LinkedIn posts until the sequence ends.

Use cases

• SaaS teams running LinkedIn-informed cold outreach across multi-touch sequences
• Automation engineers building a reliable email sequencing system with sending-limit safeguards
• Operators who want Qdrant-grounded personalization to reduce generic messaging

Technical details

• Integrations: OpenAI (Chat Model nodes), LinkedIn via RapidAPI (professional-network-data), Qdrant, NocoDB, “Send Emails” sub-workflow
• Workflow components: if, code, wait, merge, split out, aggregate
• Setup requirements: add OpenAI credentials, add Qdrant credentials and ensure an existing collection with embeddings compatible with your OpenAI embeddings model, and add RapidAPI credentials for the LinkedIn endpoints