Amazon's New AI Workspace Keeps Your Company Data In-House

Amazon and Posit have partnered so that Positron — Posit's workspace for data scientists, the app where they write code and analyze numbers — now runs inside Amazon SageMaker AI, Amazon's cloud service built for data work. Before, analysts hopped between separate tools: one for pulling company data, another for R analysis, another for Python models, another for deploying, another for reports. Now all of that happens in a single browser tab.

The practical win is privacy and less setup. The AI assistant built into Positron can use Amazon Bedrock (Amazon's service for running AI models) as its engine, so the AI help runs on models inside the company's own Amazon account and region. Amazon says customer content is encrypted, is not used to improve the underlying models, and is not shared with model providers. Data access also skips stored passwords: the workspace inherits permissions the company sets, so there are no login keys to save or rotate.

There is a catch. This isn't something an individual signs up for on a whim. It takes a company with a Posit license, an IT administrator to build and register a custom software image, and Amazon cloud accounts configured in advance. The walkthrough Amazon published uses a fake 50,000-loan dataset, so the numbers shown are a demo, not a real lending system. And it's built for people who already write R or Python code — not for casual spreadsheet users.

Still, the direction matters. AI coding assistants are becoming standard equipment, and the fight is now about where that AI runs. Vendors increasingly sell the promise that your company's data never leaves your own cloud account. For anyone whose job touches sensitive information — finance, health, customer records — that promise is often the difference between getting AI help and being told no.