canonical: https://jentic.com/apis/amazonaws.com/aws-databrew

# AWS Glue DataBrew

Jentic publishes the only available OpenAPI specification for AWS Glue DataBrew, keeping it validated and agent-ready. AWS Glue DataBrew is a visual data preparation service for analysts and data scientists. The API manages datasets, projects, recipes (transformations), recipe jobs, profile jobs, schedules, and rulesets that enforce data quality. Recipes capture hundreds of cleaning, transformation, and feature-engineering steps as code, and jobs apply them at scale to data in S3, Redshift, Snowflake, and JDBC sources.

## For AI agents

Programmatically manage DataBrew datasets, recipes, projects, and jobs to clean and transform data at scale.

## Scope

Does not handle ad-hoc SQL querying, real-time streaming, or warehouse modeling - use AWS Glue DataBrew for visual recipe-based data preparation only.

## Capabilities

- Create datasets that point to S3, Redshift, JDBC, or Data Catalog sources
- Author and version transformation recipes with hundreds of built-in steps
- Run recipe jobs and profile jobs to apply transformations and assess data quality
- Schedule recurring jobs and manage their concurrency and outputs
- Validate data with rulesets that codify data-quality assertions
- Tag, list, and delete projects, datasets, jobs, and recipes for housekeeping

## Use cases

### Self-Service Data Cleaning for Analysts

Analysts need to clean and reshape datasets without writing Spark code. AWS Glue DataBrew lets them build a recipe visually, then save it via PublishRecipe and run it programmatically with CreateRecipeJob and StartJobRun. Recipes capture every step as code, making the work reproducible and reviewable.

Example prompt: Call CreateRecipeJob with DatasetName, RecipeReference {Name, RecipeVersion}, and an S3 output, then StartJobRun and poll DescribeJobRun until the run completes.

### Profile Jobs for Data Quality Assessment

Data engineering teams profile incoming datasets to detect schema drift, null spikes, and distribution changes before downstream pipelines consume them. AWS Glue DataBrew's profile jobs sample data, compute statistics, and emit reports that integrate with rulesets for pass/fail data-quality gates.

Example prompt: Call CreateProfileJob with the dataset and an S3 output location, StartJobRun, then DescribeJobRun until COMPLETED to retrieve the profile statistics.

### Scheduled Daily Data Preparation

ETL operations need recurring transformation jobs that prepare data for analytics every day. AWS Glue DataBrew's CreateSchedule attaches a cron expression to one or more jobs, and the service handles concurrency, retries, and notifications.

Example prompt: Call CreateSchedule with a CronExpression like 'cron(0 6 * * ? *)' and JobNames=['daily-clean'] to run the job every morning at 06:00 UTC.

### Agent-Driven Data Quality Gates

An AI agent supporting a data team can run profile jobs, compare statistics against thresholds, and trigger downstream workflows. Through Jentic, the agent loads each operation by intent and chains CreateProfileJob, StartJobRun, DescribeJobRun, and CreateRuleset without writing boto3 plumbing.

Example prompt: Run a profile job, then call CreateRuleset with rules asserting completeness > 99% on key columns, and re-run the profile to validate.

## Key endpoints

| Method | Path | Description |
| --- | --- | --- |
| POST | /datasets | Create a dataset |
| POST | /projects | Create a project binding a dataset to a recipe |
| POST | /recipes | Create a recipe |
| POST | /recipeJobs | Create a recipe job |
| POST | /profileJobs | Create a profile job |
| POST | /schedules | Create a schedule |
| POST | /rulesets | Create a data-quality ruleset |

## Key resources

- **Dataset** — Create, update, list, and delete datasets that reference S3, Redshift, JDBC, or Data Catalog sources.
- **Project** — Create projects that bind a dataset to a recipe for visual editing.
- **Recipe** — Author, version, publish, and delete transformation recipes.
- **RecipeJob** — Run recipes against data sources via CreateRecipeJob and StartJobRun.
- **ProfileJob** — Run profile jobs to compute data quality statistics.
- **Schedule** — Schedule recurring jobs with cron expressions.
- **Ruleset** — Define data-quality assertions and bind them to profile jobs.

## Why Jentic

- **Setup:** Wiring AWS Glue DataBrew by hand means implementing AWS Signature v4 request signing, resolving the regional databrew host, and mapping dozens of dataset, project, recipe, and job operations to their REST paths yourself. Through Jentic you install once, import AWS Glue DataBrew from the API Directory, store the AWS credentials once, and your agent calls it.
- **Permission scoping:** DataBrew names datasets, projects, recipes, and jobs in the request body when you create them rather than pinning them in the URL path. Limit the agent to the operations it needs, such as CreateRecipeJob and CreateSchedule, so operations like DeleteRecipe are not included unless you add them.
- **Credential handling:** Your AWS credentials are stored once, encrypted, by your own Jentic One instance and injected at execution time. They never enter the agent's prompt, logs, or context.
- **Discovery method:** Agents search Jentic by intent such as 'schedule a recurring data preparation job' or 'create a data-quality ruleset', and Jentic returns the matching AWS Glue DataBrew operation with its input schema so the agent calls the right endpoint without browsing the AWS service reference.

## Related APIs

- **Amazon Athena** — Serverless SQL over data in S3, including DataBrew outputs and Data Exchange exports.
- **Amazon Athena** — Query the cleaned data DataBrew writes to S3.
- **Snowflake API** — Cloud data platform with its own visual prep tooling and marketplace.

## FAQ

### Why is there no official OpenAPI spec for AWS Glue DataBrew?

AWS does not publish an OpenAPI specification. Jentic generates and maintains this spec so that AI agents and developers can call AWS Glue DataBrew via structured tooling. It is validated against the live API and kept up to date. Get started with Jentic One, the self-hosted execution layer.

### What authentication does the AWS Glue DataBrew API use?

AWS Signature v4 (HMAC) with IAM permissions on databrew:* and pass-through permissions on the S3 buckets, Redshift clusters, or JDBC endpoints the dataset points to. Jentic stores the AWS credentials in its vault and signs each request.

### Can I run a DataBrew recipe job programmatically?

Yes. Call CreateRecipeJob with the dataset, recipe reference, and output location, then StartJobRun. Track progress with DescribeJobRun and ListJobRuns; the run completes asynchronously.

### What are the rate limits for the AWS Glue DataBrew API?

Per-account, per-region TPS limits apply, and there is a soft cap on concurrent job runs per account. StartJobRun and the Describe* operations are higher TPS than create operations.

### How do I schedule a daily DataBrew job through Jentic?

Search Jentic for 'schedule a recurring DataBrew job', load the CreateSchedule schema, and execute it with a CronExpression and the JobNames to run. Schedules apply to one or more existing jobs.

### What sources does AWS Glue DataBrew support?

DataBrew reads from Amazon S3 (CSV, JSON, Parquet, Excel), AWS Glue Data Catalog (S3, Redshift, JDBC), Snowflake (via JDBC), and Amazon Redshift directly. Outputs land in S3 in the chosen format.

### Can I limit what my agent is allowed to do with the AWS Glue DataBrew API?

Yes. Because Jentic One is self-hosted, you decide which DataBrew operations your agent may call and which AWS credentials it uses. You can allow only the operations the agent needs, such as CreateRecipeJob and CreateSchedule, so write operations like DeleteRecipe are excluded unless you add them. DataBrew names datasets, projects, recipes, and jobs in the request body rather than the URL path, so scoping happens at the operation level under your own rules.