What is SageMaker Studio?
SageMaker Studio is a secure, cloud-based workspace provided by Amazon Web Services. Instead of spreadsheet formulas, you write simple Python instructions to work with data and AI models.
You do not need to install any software. You do not need to configure anything. Your workspace is already set up and waiting for you when your sandbox access is provisioned. Everything runs in Kearney's AWS account, encrypted with a dedicated security key, and isolated from any client or production environment.
That is fine. The sandbox team provides starter notebooks: pre-written templates you can modify and run without writing code from scratch. They work similarly to pre-built Excel templates where you fill in the relevant values.
What can you do with it?
Analyze synthetic federal financial data
Run analysis against realistic synthetic records that mimic PBIS and STARS-FL financial system formats. Test detection approaches without touching real data.
Use AskSage from a compliant environment
Call the AskSage IL5 API directly from a notebook. The API key is pre-loaded. Write a prompt in Python and get a response from GPT-4o or o-series models.
Run anomaly detection experiments
Use pre-built anomaly detection models to score synthetic financial records and identify unusual patterns. Results are saved automatically to your S3 storage.
Prototype AI-assisted workflows
Build and test end-to-end workflows that combine data processing and AI analysis. Prove an approach here before it goes through formal accreditation.
How to log in for the first time
Sign into the AWS Console
Go to console.aws.amazon.com and sign in with the credentials delivered to you during provisioning. On first login you will be required to set a new password and configure multi-factor authentication (MFA) using an authenticator app on your phone.
Navigate to SageMaker
In the AWS Console search bar at the top of the page, type SageMaker and click the result. On the SageMaker home page, look for Studio in the left sidebar and click it.
Open your Studio workspace
You will see your user profile listed. Click Open Studio. The Studio interface will load in your browser. This may take 60 to 90 seconds the first time. You will see a file browser on the left and a welcome screen in the center.
Open a starter notebook
Ask your team lead for the starter notebook templates. These are .ipynb files that include working examples for common tasks: calling AskSage, loading synthetic data, and running an experiment. Upload a template using the file browser, open it, and run each cell in order using the Run button or pressing Shift+Enter.
A Jupyter notebook is divided into blocks called "cells." Each cell contains a small piece of code or text. You run cells one at a time, in order from top to bottom. When a cell runs successfully, a number appears in the brackets to its left. If it fails, an error message appears below the cell explaining what went wrong. Most errors are small typos or missing configuration values and are straightforward to fix.
Important rules for the sandbox
Use only approved synthetic data
The sandbox is authorized for synthetic and publicly available data only. Do not upload real client records, Kearney proprietary data, PII, or anything classified. If you are unsure whether a dataset is permitted, stop and ask before uploading.
Your work is not private
Everything in the sandbox is logged in AWS CloudTrail and retained for 90 days. This includes file uploads, API calls, and notebook executions. This is an internal environment, not a personal one. Do not store anything you would not want your team lead to see.
Do not share your credentials
Your AWS access key and console password are personal and non-transferable. Do not share them with colleagues. Each team member requires individual credentials. Request access through the normal provisioning process.
Contact sandbox-support@kearneyco.com. The sandbox team can provide starter notebooks, help troubleshoot login issues, and answer questions about what the environment supports.