Learning Outcomes
- Understand what Cortex is and how it fits into Foundry Connect.
- Identify which AI platforms Cortex supports.
- Learn who Cortex is designed for.
Cortex
Cortex is Foundry Connect's AI layer. It lets you connect to AI platforms, call models, and generate reports directly within your workflow, without switching between tools or platforms.
What Can You Do With Cortex?
With Cortex, you can integrate AI processes directly into your pipelines. Currently supported platforms:
| Platform | Description | Status |
|---|---|---|
| Vertex AI | Google Cloud's managed AI platform. | Available |
| Tamarind AI | Specialized models for biological data. | Available |
| Amazon SageMaker | AWS managed machine learning platform. | Coming soon |
Who Is Cortex For?
Cortex is for researchers and data scientists who want to go from raw pipeline output to AI-powered insights in one place. No separate tooling. No switching between platforms.
Prerequisites
Before using Cortex, you need to connect your AI platform credentials to Foundry Connect. This is a one-time setup per platform.
Tamarind AI
- Go to Profile > Integration Tokens
- Click New Token and select Tamarind as the platform
- Enter your Tamarind API key and save
Vertex AI
Foundry Connect supports two ways to connect your Vertex AI account.
Option 1: Service Account JSON
If you have a GCP service account credentials file:
- Go to Profile > Credentials
- Upload your service account JSON file
- Save your credentials
Option 2: GCP Cloud Integration
If you do not have a service account JSON file, Foundry Connect can integrate directly with your GCP account. Contact the support team to get this set up.