A collection of n8n workflows that leverage OpenAI to automate GitHub issue triage, customer support email handling, and supply chain incident response.
Automatically classifies and summarizes GitHub issues, then posts the results back as a comment.
- Fetch issues — Retrieves open issues from a configured GitHub repository.
- Classify severity — An OpenAI-powered Text Classifier categorizes each issue into one of five severity levels: Critical, High, Medium, Minor, or Miscellaneous.
- Summarize — A Summarization Chain (also backed by OpenAI) produces a concise summary of the issue.
- Comment — The classification and summary are posted as a comment on the original GitHub issue.
| Node | Purpose |
|---|---|
| Get issues of a repository | Pulls issues from GitHub |
| Edit Fields | Prepares issue data for classification |
| Text Classifier | Classifies issue severity via OpenAI |
| Summarization Chain | Generates a concise issue summary via OpenAI |
| Create a comment on an issue | Posts the result back to GitHub |
Monitors a Gmail inbox, classifies incoming support emails by category, generates AI-drafted responses, and routes notifications through Telegram.
- Gmail Trigger — Watches for new incoming emails.
- Classify category — An OpenAI model categorizes the email (e.g., Shipping, Technical, General).
- Enrich context — For shipping-related emails, the workflow fetches live shipping info from an external API. All branches strip and normalize the relevant data before merging.
- Generate response — An AI node drafts an appropriate email response based on the category and enriched context.
- Notify & respond — A support message is sent to a Telegram channel. The workflow then replies to the customer's email thread and sends either a Confirmation or Decline message on Telegram depending on the outcome.
| Node | Purpose |
|---|---|
| Gmail Trigger | Listens for new support emails |
| Classify Email Category | Routes emails by type via OpenAI |
| Get Shipping Info | Fetches shipping data from an external API |
| Strip Shipping Data / Strip Email Data | Normalizes data for downstream processing |
| Merge | Combines enriched data branches |
| Generate Email Response | Drafts a reply using AI |
| Send Support Message | Notifies the team via Telegram |
| Respond to Customer Email | Sends the AI-drafted reply to the customer |
| Confirmation / Decline Message | Posts final status to Telegram |
Receives shipment incident data via webhook, uses AI to assess impact and risk, generates a structured action plan, and records the outcome in Airtable.
- Webhook — Triggered by an incoming HTTP request with shipment data (e.g. shipment ID, supplier, delay hours, inventory level, customer priority).
- Analyse shipment data — An OpenAI chain analyzes the payload and returns structured outputs: impact severity, stockout risk, alternative supplier logic, and priority routing.
- Generate action plan — A second OpenAI chain produces a structured response: action plan, escalation level, and cost vs. delay tradeoff.
- Strip fields — Cleans and normalizes the data for the final step.
- Create a record — Writes the analysis and action plan to an Airtable base (e.g. “Delay Action Plan”) for tracking and follow-up.
| Node | Purpose |
|---|---|
| Webhook | Receives shipment/incident payload (GET) |
| Edit Fields | Maps incoming body for analysis |
| Analyse Shipment Data | AI analysis with structured output (severity, risk, routing) |
| Generate Action Plan | AI-generated action plan with structured output |
| Strip Fields | Prepares payload for record creation |
| Create a record | Saves result to Airtable |