We’re excited to share the latest updates and announcements designed to improve your experience with our products. This month’s issue includes:

• Announcing the Fabricate Data Agent private preview
• Structural’s AI-assisted Custom Categorical generator
• Expanded Data Vending Machine workflows in Structural
• Model-based Custom Entities in Textual
• Textual’s private preview in Microsoft Fabric

‍

Announcing the Fabricate Data Agent private preview

TL;DR: We built an AI agent that builds synthetic data for you, and we’re inviting you to join the private preview.

Be among the first to generate data at scale for any domain using agentic AI. Tonic Fabricate’s Data Agent lets you converse your way to custom synthetic data, leveraging the vast domain expertise of LLMs to achieve unprecedented realism in a matter of minutes.

• Chat to describe your ideal dataset.
• Generate and iterate on the fly via the chat.
• Export datasets as SQL, JSON, PDF, DOCX, EML files, and more.
• No manual setup. No scripts. Just the data you need.

Request access to the private preview today. Then join the official Fabricate Data Agent launch event on November 19, 2025, at 9am PT / 12pm ET to see a live chat between the Data Agent and its creator Mark Brocato, Head of Engineering for Fabricate at Tonic.ai. Working with the Data Agent, Mark will build a variety of complex, realistic datasets on the fly in real time.

With the Data Agent’s scalable, synthetic data, you and your team are free to innovate, unblocking software development, optimizing model training, and turbocharging your time-to-market.

‍

Custom Categorical generator now enhanced with AI

De-identifying sensitive categorical data in your databases is now faster than ever thanks to newly-integrated AI capabilities in the Custom Categorical generator. This update enables you to create lists of items in a flash simply by entering a prompt like “100 common pharmaceuticals” or “IATA airport codes”.

Ready to speed up your data generation workflow? Get started today by checking out our docs.

Expanded Data Vending Machine workflows

Structural has added new workflows to its Data Vending Machine to streamline your work creating unit and integration tests. Save your developers time and achieve better test coverage, with the following wins for efficiency:

• Self-service subsets: Automatically create tailored de-identified subsets based on your developers’ specific requests.
• Self-service updates to shared databases: Like the above, except the dataset is added to a shared database that mimics the current flow your developers use for testing.
• Integrating Structural data with LLM coding agents: Leverage LLM agents to create high quality integration tests using data generated with Structural.
• Self-service ephemeral datasets: Quickly and easily spin up test datasets from a library of dataset images.
• Structural chat: Chat with your de-identified data via an AI chatbot to perform data analytics without needing SQL or database knowledge.

Connect with our team to learn more and get these workflows up and running for your team.

‍

Coming soon: Model-based Custom Entities

In the next few weeks, we’ll release an exciting new feature within the Textual experience: Model-Based Custom Entities, a self-serve feature that lets you train your own entity detection models on your data, within your environment. Using LLM-powered annotation and a simple review workflow, you can create high-accuracy models tailored to your organization’s unique terminology—no data science team required. Faster onboarding, higher precision, and total data privacy, all in one streamlined experience. Reach out to our team to learn more.

Join the Microsoft Fabric private preview

Tonic Textual is coming to Microsoft Fabric! Now in private preview, you can de-identify, redact, and synthesize sensitive text data directly inside your Fabric environment—empowering privacy-first AI and analytics at scale. If you’re currently on Fabric or are hosting data in Microsoft OneLake, you can request access to the private preview here.

Small updates; big impacts

Often it’s the little things that matter most. Here’s a round up of our smaller releases.

Tonic Fabricate

• Fabricate’s Professional plan has been renamed to Fabricate Plus to reflect the added value it offers. For the up-to-date license plan information, visit our docs.
• Rule-based data generation in Fabricate can now generate a MongoDB database.

Tonic Structural

• For jobs that generate diagnostic logs, the job details page now includes an option to send the logs directly to Tonic.ai. Faster logs to us = faster debugging for you.
• We have unconditionally enhanced the Conditional generator:
  • the Conditional generator now supports non-integer numbers in conditions, and
  • for primary key columns, conditions in the Conditional generator can now refer to other columns. 🔗
• Structural’s subsetter has a new setting TONIC_SUBSETTING_MAX_DEPENDENCY_BREAK_ITERATIONS to allow you to configure the maximum number of iterations to attempt when breaking dependency cycles during subsetting preprocessing. Ideal for fine-tuning the subsetting time/integrity balance.

Tonic Textual

• The pipelines feature was sunset on October 1, 2025, as the pipelines functionality is incorporated into the datasets workflows in Textual. #bettertogether

As always, we’d love to hear your feedback on our products. What do you need? What do you love? What could be better? Send us a note at [email protected] or book time directly with our team. And for all the latest updates, be sure to check out our complete release notes.