🍁 ❄️ End of Fall 2025 Platform Update: Connector Copilot, Dynamic Dates, and Smarter Pipelines

Hi there,

The final Nexadata release of the Fall 2025 season delivers major platform improvements. At the center of this release is the new Connector Copilot, which enables Nexadata to connect to virtually any system with a REST API. This update also includes new support for dynamic date and time insertion, expanded coverage of natural language transformations, and improved data cleanup tools.

🔗 Nexadata Connector Copilot (Beta)

The Connector Copilot allows users to create datasets from any REST API using natural language, with no custom code or manual connector setup required.

Powered by agentic AI, the Connector Copilot autonomously interprets an API specification (such as an OpenAPI or Swagger file), determines how to authenticate, and constructs valid API requests that return structured, analytics-ready data. It automatically handles schema parsing, pagination, and nested fields so users can focus on describing what data they want, not how to retrieve it.

How It Works

Users start by providing an API specification. From there, the Connector Copilot will:

• Perform authentication discovery and apply the correct method
• Read the schema and data structure from the spec
• Prompt the user for a natural language description of the dataset they want
• Return a fully configured dataset that can be used directly in Nexadata Pipelines

Supported Authentication Methods

The Connector Copilot automatically detects the authentication method based on the API specification or endpoint behavior. It currently supports:

• OAuth 2.0
• API Keys
• Basic Authentication

Once detected, the Connector Copilot guides users through the credential setup without requiring technical configuration.

Current Scope and What’s Next

🔧 Available now: Create datasets from APIs using OpenAPI specs and natural language.

🚀 Coming soon:

1. Metadata-aware prompting to allow a user reference details of an endpoint, such as properties of an object.
2. Support for Pipeline Outputs, allowing Nexadata to write data back to external systems.

📅 Dynamically Insert Dates and Times

You can now insert dynamic, runtime-generated date and time values into fields using the @date function.

This works in most fields, including column names, values, and labels. It's useful for:

• Timestamping workflow outputs
• Dynamic labeling and versioning
• Automating date-based filtering

Just start typing @date and choose a format like %y-%m-%d, or type your own.

📄 View full documentation →

💬 Smarter Natural Language for Pipelines

Nexadata now supports all transformations when using natural language to add steps in a pipeline. This includes advanced operations like joining datasets, formatting values, cleaning columns, and more.

The system better understands user intent and automatically applies the appropriate logic based on the prompt's language.

Here are some example prompts you can try:

• “Join customer orders with product details using product ID”
• “Insert a column with today’s date in the format of yyyy-mm-dd”
• “Replace all invisible characters in the ‘Comments’ column”

🧹 Better Handling of Invisible Characters

Invisible characters, such as tabs, extra spaces, and blank values, can now be identified and removed or replaced using Find & Replace.

This helps clean up inconsistent or unstructured data, especially in columns where formatting issues might otherwise be hard to detect.

🔢 Find & Replace on Numeric Columns

Find & Replace now works on numeric columns. When used, the column is automatically converted to text format.

💡 If you need to perform math operations like scaling, filtering, or aggregating, use Nexadata’s Math functions instead to preserve numeric typing.

Thanks for being part of the Nexadata journey. This update sets the foundation for deeper integration and automation capabilities coming in Winter 2026.

— The Nexadata Team