aws / aws/amazon-sagemaker-feedback

[FR]: Support s3-like triggers to redirect SageMaker Pipeline success and error events to other services such as Lambda, SNS, SQS.

Open
#155 8 comments 0 reactions 0 assignees View on GitHub
bug
Dominant language
No language data
Stars
10
Forks
3
PR merge metrics
No merged PRs in 30d

Description

### Product Version

- [X] Amazon SageMaker Studio Classic
- [X] Amazon SageMaker Studio
- [X] Issue is not related to SageMaker Studio

### Issue Description

I’m a bit confused, to why the Sagemaker Pipelines do not trigger notifications as easily as S3 events. With S3, it’s straightforward to redirect notifications to an SQS queue, Lambda, or other resources. If similar simplicity were applied for triggering actions from SageMaker Pipelines, it would significantly reduce friction.

This ease of use is particularly important for professionals focused on getting things done efficiently. For example, platforms like Databricks make setting up notifications extremely intuitive. I understand that AWS prioritizes flexibility, but integrating triggers wouldn’t compromise that—it would actually enhance flexibility while offering a much-needed productivity boost for data science (DS) teams.

Being able to receive notifications is essential for any serious MLOps application. Without streamlined notification capabilities, this could become a deal-breaker when deciding whether to adopt SageMaker. Small- to mid-size companies, in particular, often trial these tools with the goal of improving DS team productivity. If the team has to spend significant effort just to get something as basic as a failure email, or even hire a person who does this, the tool might fail to deliver on its promise of increased efficiency.

### Expected Behavior

With these events we can easily pass them to a lambda function, sqs queue or even an SNS topic which send us emails. I tried event bridge and is really difficult to make something work.

### Observed Behavior

I dont get why this is not supported already.

### Product Category

Pipelines

### Feedback Category

Error Handling and Debugging

### Other Details

_No response_

Contributor guide

Open the contributing guide

Research direction

The issue names SageMaker Pipelines and requests triggers for Lambda, SQS, and SNS, but provides no repository files, tests, or implementation entry point. Start by reviewing SageMaker Pipeline event capabilities and the existing EventBridge approach described in the issue. Done would require a defined, supported way to route success and error events to the requested services.

Written by the indexing model from the issue text.

Assessment

Tech stack
aws
Domain
cloud, machine-learning
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.