n8n AI Support Automation

via Freelancer ·

Budget / Salary$250–750
TypeFreelance project
LocationRemote
Posted27 minutes ago
I need a robust n8n workflow that lets two large-language-model agents—Anthropic Claude and Google Gemini—work together to answer routine customer questions that arrive by email. The flow should:

• Ingest new support emails (IMAP or webhook) and pass the text to Claude and Gemini.
• Combine or rank their responses, then send the best draft back via email.
• If neither model reaches a high-confidence threshold, automatically reply asking the customer for more details rather than escalating to a human.
• Log every exchange on my calendar/email platform so I can track turnaround times and see threaded conversations at a glance.

Key requirements
• All logic must run in n8n; no ad-hoc scripts outside the workflow.
• Model routing, confidence scoring, and the clarification loop need to be clearly documented in the n8n nodes so I can tweak thresholds myself later.
• Use native n8n credentials management for API keys; do not hard-code secrets.
• Provide a short Loom or similar walkthrough plus the exported .json workflow so I can import it directly.

Nice to have (not mandatory): hooks for future CRM or chat integrations so I can expand beyond email later.

If you’ve already wired Claude or Gemini into n8n, this should be a straightforward build. I’ll test the flow on my sandbox mailbox; once it can handle a sample set of real support queries end-to-end, the project is complete.
php javascript c# programming software architecture api large language model n8n workflow automation
Apply on Freelancer →

Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.