AWS Comprehend n8n integrationAutomate AWS Comprehend with n8n.
Reading every message that lands in your tools stops scaling around the second hundred. The AWS Comprehend n8n integration hands that reading to Amazon's text models: 3 operations on a single Text resource, no trigger node, so whatever step runs before it decides what gets analyzed. For teams sitting on unread text.
Verified Trustpilot reviews · AI, automation & growth agency
What does the AWS Comprehend n8n integration actually do?
The AWS Comprehend n8n integration lets a workflow send a piece of text to AWS Comprehend and get a structured answer back, item after item. The node ships with n8n, so there is nothing to install and nothing extra to pay on the n8n side, on n8n Cloud or on a self-hosted instance. It exposes one resource, Text, with 3 operations: identify the dominant language, inspect the text for named entities, analyze its sentiment.
Take a support inbox. A Gmail step pulls the new messages, Detect Sentiment scores each one, and an IF node pushes anything negative into a Slack channel where somebody picks it up the same morning. The rest keeps flowing into the normal queue. Nobody reads the same thread twice.
Second case: forms that arrive in several languages. Detect Dominant Language runs first, its answer fills the Language Code of the next call through an expression (n8n's way of reading a field from the previous step, written {{ $json.field }}), and the routing that follows stops assuming everything is English.
Third: long text you want as columns. Detect Entities returns the named entities it found in a contract, a call note or a product review, and a Postgres insert turns them into rows you can query later. The catalogue lists one shared template built along those lines, workflow 965, which analyzes feedback and posts the result to a Mattermost channel.
The node covers 3 operations on that Text resource. For anything else in the Comprehend API, the HTTP Request node calls the endpoint directly and reuses the same AWS credential (the login you store once in n8n) through predefined authentication. That is the normal route on this node, not a hack.
One thing to plan for before you build: this integration has no trigger node. A workflow around it starts with a Schedule Trigger on a fixed interval, with an n8n Webhook, a URL another service calls when something happens on its side, or with the trigger of another tool entirely. If you are still weighing platforms before committing, the n8n review covers that ground.
How do you connect AWS Comprehend to n8n?
- 01
Create the AWS access key
Log in to the IAM console, open your user name in the upper right navigation bar and select Security credentials. In the Access keys section, select Create access key, choose your use case (pick Other if AWS does not prompt you), then set a description tag such as
n8n integrationso you can recognize the key months later. Reveal the Access Key ID and the Secret Access Key on the confirmation screen, because the secret is shown there. - 02
Store it as an AWS (IAM) credential
In n8n, open the Credentials menu and create an AWS (IAM) credential, the type that authenticates with an access key. Select your AWS Region first, then paste the Access Key ID and the Secret Access Key. You create this once: every workflow that touches Comprehend afterwards just picks the credential from the node's dropdown, and no key is ever typed inside a node.
- 03
Or assume a role instead of storing keys
n8n offers a second type, AWS (Assume Role), which authenticates by assuming an IAM role through AWS STS. The credentials used for the
AssumeRolecall are entered manually, or read automatically from the environment n8n runs in. That second path is the one to take if you self-host n8n on AWS infrastructure (EKS, ECS or EC2) and want to avoid keeping static keys inside n8n at all.
What can the AWS Comprehend node do?
The AWS Comprehend node exposes 3 operations across 1 resources. For each one: the node as you configure it in n8n, the required fields, and our field notes.
Text
3 operationsIdentify the dominant language
text.detectDominantLanguageWhat you see in n8n
Notes & use cases
Comprehend reads the text and returns the language it recognizes in it. Useful as a first pass, before anything that needs to know which language it is dealing with.
Key parameters
- Text: the text to send, usually an expression pointing at the field that carries it, like
{{ $json.text }}. - Simplify: on, n8n returns a simplified version of the response instead of the raw AWS data; switch it off when you want the full payload.
Inspect text for named entities, and returns information about them
text.detectEntitiesWhat you see in n8n
Notes & use cases
This is the one to reach for when you need the names inside a message rather than its tone. Comprehend inspects the text for named entities and returns information about each one it finds.
Key parameters
- Language Code: the language code for the text, chosen from the list (
en,fr,esand the others). - Text: the text to send, often the body of the document you just fetched.
- Endpoint Arn, under Additional Fields: the Amazon Resource Name of an endpoint associated with a custom entity recognition model, when the entities you care about are your own.
Analyze the sentiment of the text
text.detectSentimentWhat you see in n8n
Notes & use cases
Sentiment analysis turns a free-text review into something a workflow can branch on, without a human opening it first.
Key parameters
- Language Code: the language code for the text; set it explicitly, or wire it from a Detect Dominant Language step placed earlier.
- Text: the text to send, one item per execution of the node, so a batch of reviews goes through as a batch of items.
Need help automating AWS Comprehend with n8n?
A person reads every message.
