Ollama Email Categorization Flow

Anyone with an Outlook inbox who wants to automatically categorize and sort emails using AI. This workflow uses AI to read incoming Outlook emails, categorize them (e.g., junk, receipt, business), and then updates the email with the assigned category. It can also move the categorized emails into specific Outlook folders. Set up an n8n instance, configure Outlook and Ollama credentials, and import the workflow. Adjust categories and target folders as needed in the "varID & Category1" and Switch nodes. n8n, Microsoft Outlook account, Ollama API access with a Qwen2.5:14b model. You can customize the categories in the "varID & Category1" node and the AI prompt node. You can also change the target Outlook folders for moving emails. Key Technologies: n8n,Microsoft Outlook,Ollama (AI) Automation Value: - Automatically categorizes and sorts emails, saving manual effort. - Improves email organization by assigning relevant categories. - Helps in managing email overload by identifying and moving junk or less important emails. - Facilitates quicker processing of important emails by grouping them into specific categories. Best Practices: - Clean Email Data: Always sanitize email content before feeding it to an AI model to improve accuracy and reduce token usage. - Clear AI Prompting: Provide clear instructions, examples, and constraints (like allowed categories and output format) to the AI for more consistent and accurate results. - Error Handling: Implement error handling (e.g., the onError setting and Catch Errors node) for AI output to ensure workflow resilience. - Granular Categorization: Use both main categories and subcategories to provide more detailed organization. - Filtering: Filter out already processed items (flagged or categorized emails) to avoid redundant work and API calls. - Looping for Batches: Use "Split In Batches" for individual item processing, especially when external API calls or complex logic is involved for each item. - Conditional Logic: Use "Switch" and "If" nodes to create dynamic workflows based on AI output or other email properties.