1.0.14 Release Note_September 07, 2026
Release Date: September 07, 2026
The Business AI Cloud 1.0.14 release introduces the following new or enhanced features.
Important
This document provides high-level information on the latest release features and enhancements for the Business AI Cloud. For detailed procedures, configuration steps, and comprehensive guidance, please refer to the Business AI Cloud User Guide.
Release Highlights
New Features
Job Hooks - Automates custom pre- and post-processing actions around pipeline runs in a Knowledge Cluster (KC), eliminating manual script execution.
SQL Parameters - Introduces reusable, overridable parameters for SQL queries in pipelines and dataflows, removing the need for hardcoded values.
Process Intelligence (PI) - Adds BAIC Copilot support for Process Explorer, Process Occurrences, and SOP Studio within Process Intelligence, letting you view process data quickly and update an SOP using natural language instead of manually editing the process map.
Agent Guardrails - Introduces automated safety checks that scan every agent interaction for prompt-injection attempts and sensitive information (PII/secrets), with results visible in Agent Chat and Agent Evaluation.
Custom App Generation - Share link - Creates a shareable, no-login public URL for agent-created custom apps, with expiration and revocation controls.
Enhancements
Configuration Export/Import - Exports and imports Knowledge Base, Knowledge Cluster, and Dataflow configurations, simplifying reuse across projects.
Data Orchestrator Agent - Adds the ability to modify configuration during chat-based dataflow creation, before approving the generated dataflow.
Data Pipeline Mapping Export - Downloads pipeline mapping to your local system in CSV or Excel format.
Process Intelligence - Adds an interactive node graph (Context Graph) to Process Explorer that shows how process steps, entities, and processes relate through connections.
Hybrid Search for Knowledge Bases - Combines Agentic Search to narrow retrieval to the most relevant files using the Knowledge Graph, and Semantic search to then find the precise answer within those files.
LLM/SLM Management - Redesigns the page to provide a cleaner, easier-to-use interface.
Credit Actuals - Extends its token usage tracking capability to Data Agents, enabling tracking of data pipeline and dataflow LLM usage.
Agent Optimization - Fine-tunes SLMs and optimizes prompts for any agent using natural language requests to the BAIC Copilot, which analyzes the agent and takes the needed action.
Agent Chat - Redesigns the Agent chat page for a cleaner, easier-to-use interface, splitting traces and agent details from the conversation view.
New Features
Job Hooks
Business AI Cloud lets you attach custom scripts to pipeline jobs within a KC that run automatically before or after a job executes. Job Hooks remove the need to manually trigger auxiliary tasks such as setting file permissions, sending notifications, or cleaning up data each time a pipeline runs. Because a hook is defined once at the KC level, it can be reused across any pipeline built from that cluster, ensuring consistent, repeatable behavior without duplicated setup.
Key Capabilities:
Define a custom script once at the KC level and reuse it across any pipeline built from that cluster.
Attach a hook as Pre- or Post-Job Hook to control whether it runs before or after the pipeline job executes.
Hooks run automatically on every pipeline execution, removing the need for manual script runs.
The Job Hooks feature is available in Data Discovery Agent within Business AI Cloud's Data Layer.
To access, go to the Home page > Data > Data Agents > Data Discovery Agent > Knowledge Cluster.

For more information, refer to Knowledge Clusters in the User Guide.
Validation Templates and Pipeline Validation
Business AI Cloud lets you define Validation Templates that check whether the data written to a pipeline's target meets your rules, such as confirming that a column has no missing values or that an average falls within an expected range. Each template is associated with a specific Knowledge Cluster and its related pipeline. You can then apply an existing template to a specific pipeline version and define validation, resolving its parameters with real values such as table or column names. Together, these capabilities confirm that pipeline output meets the defined quality checks before agents and reports rely on it.
Key Capabilities:
Reuse the same Validation Template across multiple pipeline versions by resolving parameters differently each time.
Supports common data quality checks such as null-value detection and range validation.
Surfaces data quality issues at the pipeline level, before the data reaches downstream agents and reports.
The validation templates and pipeline validation feature are available in the Data Engineering Agent within Business AI Cloud's Data Layer.
To access validation templates, go to the Home page > Data > Data Agents > Data Engineering Agent > Validation Template.

For more information, refer to Managing Validation Templates in the User Guide.
To access pipeline validation, go to the Home page > Data > Data Agents > Data Engineering Agent > Pipeline Version > Validation.

For more information, refer to Viewing Pipeline Validation in the User Guide.
SQL Parameters
Business AI Cloud lets you define reusable SQL parameters, such as date ranges, thresholds, or filter values, instead of hardcoding them into your queries. Parameters are set once per pipeline or dataflow version and referenced directly in the SQL, keeping the query logic stable even as the underlying values change. When a run requires a different value, you can override the parameter for that single run without editing the saved version. This capability applies equally to both pipelines and dataflows, giving teams a consistent way to manage changeable values across their data workflows.
Key Capabilities:
Override parameter values for a single manual run without editing the saved version.
Separates changeable values from query logic, keeping SQL reusable and easier to maintain.
Consistent parameter behavior across both pipeline and dataflow SQL.
The SQL Parameters feature is available in the Data Agent within Business AI Cloud's Data Layer.
To access, go to the Home page > Data > Data Agents > Data Engineering Agent > Pipeline Version > SQL Editor.

For more information, refer to Adding SQL Queries in the User Guide.
To access, go to the Home page > Data > Data Agents > Data Orchestrator Agent > Dataflow.

For more information, refer to Managing a Dataflow Version in the User Guide.
Process Intelligence
Business AI Cloud adds BAIC Copilot support across Process Explorer, Process Occurrences, and SOP Studio within Process Intelligence. Instead of manually navigating each module to view process data or editing the process map directly to update an SOP, you describe what you need in natural language, and the Copilot retrieves the relevant process information or applies the SOP change. This brings a single conversational entry point to three previously separate, manually driven workflows.
Key Capabilities:
Query process data in Process Explorer and Process Occurrences using natural language, without manually navigating each module.
Update an SOP in SOP Studio using natural language, without manually editing the process map.
BAIC Copilot is available within Process Intelligence in Business AI Cloud's Data Layer.
To access Copilot from SOP Studio, go to the Home page > Data > Process Intelligence > SOP Studio.

For more information, refer to SOP Studio in the User Guide.
Agent Guardrails
Business AI Cloud automatically checks every agent interaction for unsafe or sensitive content. Two checks run in the background for each interaction: one on the user's prompt before it reaches the agent, and one on the agent's response before it reaches the user. Guardrails currently detect prompt-injection attempts and PII/secrets, such as credit card numbers, API keys, passwords, and email addresses. Results are flagged and logged for review; no content is currently blocked, modified, or redacted.
Key Capabilities:
Runs automatically on every agent call, with no configuration required.
Detects prompt-injection attempts, in which a user tries to manipulate the agent into ignoring its instructions.
Detects PII/secrets in conversation text, including credit card numbers, API keys, passwords, phone numbers, and email addresses.
Flags results per user, supporting later audit and compliance review.
Guardrails run automatically within Business AI Cloud's Agent Layer. Results are available in these places:
Agent Chat - A shield icon appears next to any message where a check was triggered.

Agent Evaluation - The Guardrails tab displays a report of guardrail outcomes, including any triggered violations or enforcement actions.

For more information, refer to Agent Guardrails in the User Guide.
Custom App Generation - Share link
Business AI Cloud lets you generate a public URL for custom apps created to display agent-generated data, so you can share the app with others without requiring them to log in to the platform. This removes the need for external viewers to create an account or log in just to see a report or dashboard built on agent output. You can set an expiration window for the link and revoke it at any time, giving you control over how long the app remains accessible and to whom.
Key Capabilities:
Set an expiration window so the link stops working automatically after a defined period.
Revoke a link at any time to immediately cut off access, independent of the expiration setting.
Useful for sharing agent outputs with external stakeholders without provisioning platform accounts.
The Public Link feature for Custom App is available within Business AI Cloud's Agent Layer.
To access, go to the Home page > Agents > Agent SDK > Custom App

For more information, refer to Exporting an App in the User Guide.
Enhancements
Configuration Export/Import
Business AI Cloud lets you export a Knowledge Base (KB), Knowledge Cluster (KC), or Dataflow configuration and import it into another KB, KC, or Dataflow. This removes the need to manually recreate similar setups from scratch each time a new project requires the same configuration. Teams working across multiple projects or environments can reuse a proven setup instead of rebuilding it field by field.
Key Capabilities:
Supported for KBs, KCs, and dataflows, covering individual, grouped, and workflow-level configurations.
Reuse proven configurations across projects and environments rather than rebuilding them.
Reduces setup time and configuration drift when replicating similar setups.
The Configuration Export/Import feature is available within Business AI Cloud's Data Layer.
For more information about KB configuration, refer to Exporting and Importing Configuration in the User Guide.
For more information about KC configuration, refer to Config Tab in the User Guide.
For more information about dataflow configuration, refer to Exporting and Importing a Dataflow in the User Guide.
Data Orchestrator Agent
Business AI Cloud enhances the Data Orchestrator Agent, letting you add or adjust configuration while creating a dataflow through the conversational chat interface, before approving it. This gives you the opportunity to refine the generated dataflow to your requirements before it's finalized, rather than approving it first and making changes afterward.
Key Capabilities:
Add or modify configuration directly during chat-based dataflow creation.
Review and adjust the generated dataflow before approval to ensure it meets the requirements up front.
This enhancement is available within the Data Orchestrator Agent in Business AI Cloud's Data Layer. For more information, refer to Creating a Dataflow with an Agent in the User Guide.
Process Intelligence
Business AI Cloud adds a Context Graph view to Process Explorer, showing how the entities, steps, and processes within a selected process relate to one another as an interactive node graph. Rather than viewing a process as a linear sequence, you can view how process steps connect to data entities and how relationships such as traverses, reads, and writes link them together. This surfaces data dependencies and structural relationships that aren't visible before. The graph supports multiple layout options and filtering by node and relationship types, allowing you to focus on the specific dependencies relevant to your analysis.
Key Capabilities:
Displays process steps and entities as a connected node graph, revealing relationships not visible in linear or analytics-based views.
Offers three layout options: Layered, Radial, and Force, to suit different graph structures and analysis needs.
Filters the graph by node type (Step, Entity, Process, Path) and relationship type to isolate specific dependencies.
The Context Graph view in Process Explorer is available within Process Intelligence in Business AI Cloud's Data Layer. For more information, refer to Process Explorer in the User Guide.
Hybrid Search for Knowledge Bases
Business AI Cloud offers Hybrid Search for Knowledge Bases, a smarter way to search that narrows down the right files first, then looks inside them for the answer. It combines Agentic Search, which uses the Knowledge Graph to identify the files most relevant to a question, with semantic search, which then finds the precise answer within those files. This two-stage approach avoids searching the entire knowledge base for every query. Instead, it focuses effort on the content most likely to contain the answer, improving both relevance and precision.
Key Capabilities:
Combines Semantic Search and Agentic Search into a single retrieval approach.
Improves answer precision for large or multi-document Knowledge Bases, where full-KB semantic search alone may surface less relevant results.
Available as a retrieval option alongside existing search methods within the KB.
This Hybrid Search is available within Business AI Cloud's Knowledge Layer. For more information, refer to Modifying a Knowledge Base in the User Guide.
Agent Optimization
Business AI Cloud lets you fine-tune the SLM and optimize prompts for any agent on the platform using the BAIC Copilot. Instead of manually configuring fine-tuning runs or rewriting prompts, you describe what you need in natural language, and Copilot analyzes the agent and acts on the request sequentially. This removes the need for manual, step-by-step optimization work and puts SLM fine-tuning and prompt optimization within reach through a single conversational interface.
Key Capabilities:
Covers both SLM fine-tuning and prompt optimization from a single conversational entry point.
The Copilot determines and executes the necessary steps based on the request, rather than requiring each configuration step to be specified manually.
BAIC Copilot for Agent Optimization is available within the Agent SDK in Business AI Cloud's Agent Layer. For more information, refer to Optimizing Agents in the User Guide.
Agent Chat
Business AI Cloud redesigned the Agent Chat page for a cleaner, easier-to-use interface for agent testing. Traces and agent details appear on the left panel, while your interactions and agent responses appear on the right panel. This layout separates conversation from supporting context, making it easier to follow agent behavior alongside the chat itself.
Key Capabilities:
View traces and agent details on the left panel, separate from the conversation.
Navigate agent behavior and conversation history side by side without switching views.
This Agent Chat is available within Business AI Cloud's Agent Layer.
![]() |
For more information, refer to Managing Agents in the User Guide.
You can also access the Agent Chat from the Home page.

Constraints and Limitations
The following constraints and limitations exist for this release:
Item | Description |
|---|---|
Agent Chat | File attachments have the following limits:
|
