AI-powered ServiceNow chat triage with GPT-4 — incident & request manager
# Automated Incident and Request Management in ServiceNow
### Who’s it for
This workflow is designed for IT teams, service desk agents, and operations managers who use ServiceNow. It reduces manual effort by automatically classifying chat messages as **Incidents** or **Requests**, creating/updating them in ServiceNow, and summarizing ticket updates.
---
### What it does
* Receives incoming chat messages.
* Classifies the message as one of:
* **Incident** (something broken, unavailable, or a complaint)
* **Request** (access, provisioning, product/order related)
* **Follow-ups** (incident or request update checks)
* **Update action** (user wants to add info to an existing ticket)
* **Everything else** (knowledge search / general query).
* Creates **Incidents** in ServiceNow via the ServiceNow node.
* Creates **Requests** in ServiceNow using the Service Catalog API.
* Updates existing Incidents with new work notes when the user provides an update.
* Pulls existing incident/request work notes for summaries.
* Optionally uses **SerpAPI** for general queries (if enabled).
* Returns a concise summary back to the user through the webhook.
---
### Requirements
* **ServiceNow account** with API access (Basic Auth)
* **OpenAI API key** (used by the classifier and summarizer)
* **SerpAPI key** *(optional – for general web lookups)*
---
### Credentials needed
You will need to set up the following credentials in n8n:
1. **ServiceNow Basic Auth** (username, password, instance URL).
2. **OpenAI API** (API key).
3. **SerpAPI** *(optional – only if you want web search enabled)*.
---
### How to set up
1. Import the workflow JSON into your n8n instance.
2. Create the credentials mentioned above and assign them to the corresponding nodes:
* `Create an incident` → **ServiceNow Basic Auth**
* `HTTP Request1` (for Service Catalog requests) → **ServiceNow Basic Auth**
* `OpenAI Chat Model` / `OpenAI Chat Model1` / `OpenAI Chat Model2` / `OpenAI Chat Model3` → **OpenAI API**
* `SerpAPI` node (optional) → **SerpAPI key**
3. Adjust the **ServiceNow instance URL** in the HTTP Request node to match your environment.
4. Deploy the workflow.
5. Send a test chat message to trigger the workflow.
---
### How to customize
* Update the **classification rules** in the **Text Classifier** node if your organization uses different definitions for incidents vs. requests.
* Edit the **summary prompt** in the Summarization Chain to include or exclude specific fields.
* Add additional notification nodes (Slack, Teams, or Email) if you want updates pushed to other channels.
---
### Notes & Limitations
* This workflow creates **general requests** in ServiceNow using the catalog API. For production, update the Service Catalog item ID to match your environment.
* “Everything else” category uses SerpAPI. If not configured, those queries will not return results.
* This workflow requires **OpenAI GPT-4.1 mini** (or another supported model) for classification and summarization.
---
