API ReferenceProduct Documentation
Product Documentation
Start typing to search…

TetraScience AI Services v1x Release Notes

The following are the release notes for TetraScience AI Services versions 1.x Tetra AI Services empower developers and scientists to transform data into decisions through seamless, secure APIs, and enable intelligent automation across scientific, lab, and enterprise workflows.

📘

NOTE

TetraScience AI Services are currently available to Scientific AI Lighthouse (SAIL) Partners only as part of a limited availability release and must be activated in coordination with TetraScience. For more information, contact your customer account leader.

v1.0.0

Release date: 4 December 2025

What's New

TetraScience has released its first limited availability version of TetraScience AI Services, version 1.0.0. Available through the Tetra Data Platform (TDP), AI Services introduces foundational capabilities for deploying and operating Scientific AI Workflows in production.

While model training happens outside of AI Services, this release provides managed compute and model-serving infrastructure so customers can run AI workflows without handling clusters, endpoints, or scaling. Each workflow artifact is versioned, access-controlled, and promoted through admin controlled activation. AI Services also captures key inference and operational metrics usage, performance, and errors to support monitoring and audit needs. This foundation enables customers to operationalize AI reliably while building toward future capabilities.

Here are the details for what's new in TetraScience AI Services v1.0.0.

Prerequisites

TetraScience AI Services v1.0.0 requires the following:

New Functionality

New functionality includes features not previously available in the Tetra Data Platform (TDP).

AI Services User Interface

📘

NOTE

The AI Services user interface doesn't load in TDP v4.4.0. A fix for this issue is scheduled for TDP v4.4.1. As a workaround, customers can use the AI Services API endpoints to get the same functionalities programmatically. For instructions on how to validate core AI Services functionality programmatically, see Test TetraScience AI Services Programmatically in the TetraConnect Hub. For access, see Access the TetraConnect Hub.

A new, dedicated AI Services interface that can be accessed within the TDP through Artifacts > AI Services provides the following:

  • Scientific AI Workflows: Browse and review workflow versions with associated metadata.

  • Activate Workflow: Admins can activate a specific workflow version to make it available for authorized users.

  • Install Workflow: Provision the required compute resources on-demand, resulting in an inference endpoint you can invoke.

  • Online Inference: The online inference endpoint invokes the appropriate LLM, AI agent, or custom model to return real-time predictions from text-based inputs.

  • Offline Inference: Provide inputs through JSON files or specify datasets for batch predictions and asynchronous processing.

  • Job Monitoring: Track failed inference jobs through the TDP Health Monitoring Dashboard.

  • Uninstall Workflow: Remove the compute resources for a workflow without deleting its artifacts or versions.

For more information, see the TetraScience AI Services User Guide (v1.0.x) .

AI Services page

Scientific AI Workflows

AI Services supports many Scientific AI Workflows for specific use cases, which will be released independently from the AI Services user interface and the TDP as their own artifacts. The scientific use cases that the AI Workflows support are developed as Embedded Data Apps in the Tetra Data & AI Workspace. Updates to all three artifacts (AI Services, AI workflows, and Data Apps) can be released independently.

For more information about how to run inferences on installed Scientific AI Workflows, see Run an Inference (AI Services v1.0.x) .

Scientific AI Workflow example

AI Services APIs

AI Services provide five core APIs that enable AI-driven tasks:

  • Submit an online (synchronous) inference request(/inference/online): Performs real-time ML inference and returns results synchronously.
  • Submit an offline (asynchronous) inference request(/inference): Submit an asynchronous, file-based inference request with support for partial file staging success. Returns a request ID and status URL for tracking progress.
  • Health check (/health): Returns a 200 response if the service is running and supports error simulation for testing.
  • Get inference request status (/inference/{inferenceId}/status): Retrieves the current status and details of an inference request. Includes file-level details and partial success information when applicable.
  • Get metrics for a synchronous (online) inference request (/inference/online/{requestId}/metrics): Retrieves metrics and status information for a completed synchronous (online) inference request. Metrics are stored for tracking and debugging purposes.

For more information about how to use these endpoints to run inferences on installed Scientific AI Workflows, see Run an Inference (AI Services v1.0.x) .

AI Workflow Lifecycle Management

AI Services provide the following AI Workflow lifecycle management capabilities:

  • AI Workflow Installation: Install a specific workflow version to provision the required compute infrastructure on Databricks (model serving endpoints, jobs depending on the workflow).
  • AI Workflow Versioning: Maintain multiple versions of an AI Workflow, each with its own packaged assets and configuration. AI Services track versions for discoverability, installation, upgrade paths, and controlled rollout across environments.
  • Multiple AI Workflow Types: Support for online inference using foundational LLM, custom models, AI Agents (synchronous), custom models (asynchronous) and tools workflow types.

Dynamic AI Agent Deployment

The platform supports dynamic deployment of AI agents that encapsulate task-specific instructions and logic:

  • Category/Alias-based Routing: Route requests to appropriate agents using categories or aliases (for example, Coding Agent, and Pipeline Agent).
  • Agent Behavior Configuration: Define agent behavior through instruction templates and knowledge base associations.

Inference Capabilities

AI Services provide the following flexible inference execution options:

  • Synchronous Inference: Quick, interactive responses for real-time use cases.
  • Asynchronous Inference: Batch processing for larger datasets with job tracking and notifications.
  • Image Upload for Inference: Upload single or multiple images from scientific instruments for analysis.
  • Pipeline Integration: Integrate inference into data pipelines using task scripts with specified endpoints.
  • GetData API: Query for output data using request UUIDs once inference is complete.

Enhancements

Enhancements are modifications to existing functionality that improve performance or usability, but don't alter the function or intended use of the system.

Platform Operations

  • Independent Releases: AI Services releases independently of TDP releases.
  • Terraform-controlled Activation: Platform activation managed through Terraform without requiring customer AWS CloudFormation templates.
  • Backward Compatibility: Newer AI Services versions maintain backward compatibility with TDP releases v4.4.0 and higher (TDP v4.4.1 or higher is recommended).
  • Dynamic Resource Scaling: Databricks compute resources scale up to meet demand, maintain for sequential requests, and scale down after 5 minutes of inactivity.

Developer Experience

  • Cookie Cutter/TS-CLI: Project setup templates for standardized AI/ML project structure.
  • YAML-based Configuration: Configure jobs, parameters, and artifacts using YAML files.
  • MLflow Integration: Templatized MLflow format for compute configuration and workflow tracking.
  • Development Environment Support: Compatible with VSCode and Databricks notebooks.

Observability and Monitoring

  • Platform-level Observability: Monitor data flow and infrastructure health.
  • Product Observability: Monitoring and troubleshooting capabilities.
  • Job Status Tracking: Comprehensive tracking of submitted jobs with success/failure reasons.
  • Email Notifications: Automatic push notifications upon job completion.

Bug Fixes

There are no bug fixes in this initial release.

Deprecated Features

There are no deprecated features in this initial release. For more information about TDP deprecations, see Tetra Product Deprecation Notices.

Known and Possible Issues

The following are known limitations of Tetra AI Services v1.0.0:

  • UI Loading Issues in TDP v4.4.0: The AI Services user interface doesn't load in TDP v4.4.0. A fix for this issue is scheduled for TDP v4.4.1. As a workaround, customers can use the AI Services API endpoints to get the same functionalities programmatically.
  • Task Script README Parsing: The platform uses task script README files to determine input configurations. Parsing might be inconsistent due to varying README file formats.
  • AI Agent Accuracy: AI-generated information cannot be guaranteed to be accurate. Agents may hallucinate or provide incorrect information.
  • Organization-Specific Components: AI capabilities are limited to publicly available TetraScience components and cannot incorporate organization-specific components.
  • Complex Logic Implementation: AI-generated pipelines with complex logic may require manual implementation and refinement.
  • Databricks Workspace Mapping: Initially, one TDP organization maps to one Databricks workspace only.
  • Analyst Policy Install Workflow UI Access: Users assigned a role with an Analyst policy can view, but not use, the Install Workflow button even though the policy doesn't allow them to install an AI workflow or run an inference. A fix for this issue is scheduled for a future AI Services UI release.

Security

AI Services extends existing TDP security practices:

  • Data Encryption: Maintains data encryption standards
  • Data Residency: Supports data residency requirements
  • Penetration Testing: Regular security assessments
  • JWT Authentication: Secure API access with JWT tokens
  • Tenant-level Authorization: Scoped access to workflows, agents, memory, and context based on JWT claims.

For more information, see Product Security.

Compliance

  • AI Clauses in Contracts: Includes disclaimers and transparency requirements for AI tool usage.
  • AI Workflow Observability: Tracking and monitoring for data and workflow drift detection.
  • Audit Trail: Comprehensive traceability for trained AI Workflows including code, hyperparameters, and training data.

Upgrade Considerations

TetraScience AI Services are currently available to Scientific AI Lighthouse (SAIL) Partners only as part of a limited availability release and must be activated in coordination with TetraScience.

The AI Services user interface doesn't load in TDP v4.4.0. A fix for this issue is scheduled for TDP v4.4.1. As a workaround, customers can use the AI Services API endpoints to get the same functionalities programmatically. For instructions on how to validate core AI Services functionality programmatically, see Test TetraScience AI Services Programmatically in the TetraConnect Hub. For access, see Access the TetraConnect Hub.

For future upgrades, AI Services will release independently of TDP releases, with activation controlled through Terraform and feature flags.

For more information, see the TetraSciences AI Services documentation.

Updated about 2 hours ago