Draft keigo-checked Japanese contact form replies with OpenAI and Teams

## Draft keigo-checked Japanese replies to contact forms from an FAQ table with OpenAI and Teams Replying to Japanese customers takes time: business emails need correct keigo, the right opening and closing, and accurate answers. This workflow drafts each reply from your own FAQ table, has a second AI pass check the keigo, and posts the result to Microsoft Teams so a person can review it before sending. ### How it works 1. A customer submits the contact form. 2. The workflow loads your FAQ entries from an n8n data table and adds your company name and signature. 3. AI Classify Inquiry sorts the message into six types. Spam is skipped, and each other type adds its own reply guidance. 4. Find Matching FAQ picks up to three FAQ entries whose keywords appear in the message. 5. Generate Keigo Reply writes a Japanese email using only those answers, rates urgency, and lists what staff must check. Questions without an FAQ answer are flagged instead of guessed. 6. Check Keigo reviews the draft for double keigo, overused させていただく, mixed-up respectful and humble forms, and casual phrases, and lists every fix. 7. The draft, the fixes, and the FAQ entries used are posted to Teams. ### Setup steps 1. Create an n8n data table named keigo_faq with three text columns: topic, keywords (comma-separated), and answer. 2. Run Load Sample FAQ once to add sample rows, then replace them with your own. 3. Set your company name and signature in Set Company Details. 4. Connect your OpenAI credential to OpenAI gpt-5-mini. 5. Paste a Teams Workflows webhook URL into Post to Teams Webhook, then share the form's production URL.