Capabilities

Platform updates

Workflow Builder

Ontology

Workflow Builder is a new application designed to help with building, debugging and managing workflows. Available in beta across all enrollments the week of November 11th, you can add Ontology resources – objects, LLMs, models, actions, functions, applications – to the graph and expand to other resources with logical references. Workflow Builder graphs are autogenerated - simply use the keyboard shortcut `CMD + i` (macOS) or `Ctrl + i` (Windows) when viewing any Workshop application to get started immediately. [Review the documentation on Workflow Builder](/docs/foundry/workflow-builder/overview/).

Introducing Workflow Builder, an interactive workspace for understanding your workflows [Beta]

Projects

Security

We are excited to announce the GA release of Project cover pages, available now on all enrollments. The Project **Cover Page** section offers a Markdown-based rich-text editor for writing comprehensive documentation about the Project.

Introducing cover pages for discoverable sensitive projects [GA]

AIP Agent Studio

[AIP Agent Studio](/docs/foundry/agent-studio/overview/) now features an **Agent details** panel to the right of the conversation window which provides insight into the chosen model, prompting strategy, available resources, retrieval context, and reasoning methods of an agent deployed in [AIP Threads](/docs/foundry/threads/overview/) or [Workshop](/docs/foundry/workshop/overview/). The panel aims to provide additional transparency to help users obtain the clarity necessary to configure and best leverage their Agent.

Explore new Agent details panel for configuration insight in AIP Agent Studio [Beta]

AIP Assist

A new integration between Code Repositories and AIP Assist will be available in a beta state the week of November 11th. This integration introduces context-aware attachments for AIP Assist that allow users to attach existing code snippets, files, and even entire repositories to conversations. This provides context that enriches AIP Assist knowledge of user workflows, allowing Assist to respond to code-specific questions more accurately and with greater detail. 

Improved developer support with context-aware attachments in Code Repositories [Beta]

An enhanced AIP Assist user experience will be generally available starting the week of November 11th. This update brings performance and usability upgrades designed to improve the overall AIP Assist experience. With this new release, you can look forward to improved product behaviors and new features, like the ability to extend the sidebar, improved keyboard navigation, and real-time synchronization across tabs.

Performance and usability improvements in AIP Assist

Code Repositories

Data Integration

Instead of modifying an example file when configuring new Python transforms in Code Repositories, you can now use our new file template configuration wizard for an immediate output that fits your needs. Use the configuration wizard by creating a new Python Transforms repository, or by selecting **Add > New file** from template using the Files side panel. This new feature facilitates the personalization of the default file in new repositories by enabling you to choose the transformation type along with necessary inputs and outputs.



Use the new file template configuration wizard in Code Repositories to create transforms with ease

Control Panel

Administration

You can now configure many Organization-wide Code Repositories settings within Control Panel. To modify these settings, you will need the `User experience administrator` role, also configurable in Control Panel under Organizations permissions.



Configure Code Repositories settings in Control Panel [GA]

By default, while using an [Action Tool](/docs/foundry/agent-studio/tools/#types-of-tools), Agents may propose that an action should be executed with specific inputs. However, user input is necessary to apply the action. You can now allow for action tools to run automatically by enabling the new **Allow agent to automatically run the selected action** toggle in the Action Tool configuration. For clarity, automated actions are also marked with a tag in the reasoning panel and in the Agent's read-only view.

Automated Action Tool

Approvals

We have removed several redundant approval statuses from the status filter dropdown menu in the Approvals application, improving and clarifying the user experience.

Remove redundant approval statuses

Palantir API

The Palantir platform is the Ontology-powered operating system for the modern enterprise, and it now comes with powerful developer tools that allow users to interact with the platform programmatically. Along with existing generally available features, APIs, and SDKs, we have released new beta functionality and will continue to expand our suite of developer tools in the coming weeks and months as part of our commitment to a development ecosystem tailored to your business and operational needs.



The Palantir platform is now a fully fledged development ecosystem

Notepad

Analytics

The custom functions feature in Notepad allows users to use Foundry-native functions to modify the content in a Notepad document. This feature has been enhanced to support Markdown for both input and output in custom functions. Previously, any selected content was converted into plain text before being passed to the function, which stripped away valuable formatting such as lists, bold, italics, and underline. This conversion resulted in the loss of important context and meaning derived from the formatting.

Parse rich text from custom Functions

You can explore all existing versions of a document by using the **Version history** panel on the left. A version is a snapshot of the current state of the document that you can revisit or even restore at a later point in time. You can also manually save versions. By default, only user-created versions will be shown. Switch the filter at the top right of the panel to **All versions** to view all autosaved and manual versions. In addition to autosave versions, you can also create manual versions.

Version history now available

Pipeline Builder

Pipeline Builder now includes a new check for pipelines on a Foundry branch. This check surfaces errors preventing the user from deploying the Foundry branch the pipeline is on. Users can view specific error details by selecting the **Details** dialog. Proposals that fail this check are not mergeable, and the user will not be able to deploy the Foundry branch.

Add proposal check for Foundry branch deployability in Pipeline Builder

Modeling

Model Integration

Modeling application users can now modify the API name and row-wise configuration for functions post-publication.

Enhanced Editing for Published Functions in Modeling

Ontology Management

You can now allow users to directly upload media to an object with Actions. Configure your Action form to take a media reference parameter, and the form will provide a clean and intuitive interface for form users to upload their media.

Directly upload media to an object with Actions

If users do not have permission to edit an Action type, a list of required `Edit` permissions will appear to guide users towards next steps to gain access to Ontology edits.

Display required permissions for Action type edits

Code Workspaces

You can now prevent Jupyter® code workspaces from becoming idle when executing long-running processes. Review [the relevant section in the FAQ](/docs/foundry/code-workspaces/code-workspaces-faq/#how-do-i-prevent-my-jupyter-workspace-from-pausing-while-my-code-is-running) for more details.

New keep_alive cell magic to prevent notebooks from becoming idle

The functionality to apply custom functions to Notepad content has been upgraded to support Markdown formatting. This enhancement ensures that formatting elements such as lists are preserved in both the input and output of functions, potentially unlocking new workflows with custom functions.

Parse rich text from custom functions via Edit with AIP

Quiver

Quiver introduced an array of enhancements to event set workflows. Users will find new functionality for generating event sets from tabular data, inputting ranges manually, conducting object search-arounds, and adjusting event timings. The chart legend/editor now includes a separate event plot section.

Enhanced Quiver event set capabilities

Slate

App Building

Slate users can now hide the navigation sidebar in application view mode by navigating to settings and toggling **Show navigation sidebar in view mode**. The sidebar will still be displayed by default in view and edit modes.

New setting allows users to hide navigation sidebar in Slate

Workshop

Workshop now automatically saves edits to your module.  If you leave or reload the page after making changes to your module, your edits will be retained as a draft. You can then choose to discard your draft on page load or from the autosave indicator in the top left corner of the application.

Automatic saving

Media sets

TIFF files are now supported within media sets, expanding the range of acceptable media formats and enhancing the flexibility of media management workflows.

TIFF files are now supported within media sets

This updates the valid character set for backing branch names to include everything except `%` and whitespace. For example, the special character of `/` and `&` are now supported.

Different characters are now usable for branch name in Pipeline Builder

Users can now dynamically resize the resource picker in the **Add data** modal of the Pipeline Builder. This enhancement allows for customizable visibility, making it easier to view and manage long dataset names when importing data.

Adjustable 'Add Data' Interface in Pipeline Builder

Ontology Manager now prevents the direct removal of inherited Structured Property Types (SPTs) from interfaces. The **Remove from interface** button is now disabled for inherited SPTs, with a tooltip providing an explanation upon hover.

Removing inherited SPTs from an interface is now disabled

Ontology Manager now provides more informative error messages for incomplete configurations involving classification properties. Specifically, it alerts users when a maximum classification is not set for a CBAC property or when allowed markings are missing on a Mandatory marking property.

Error message update for max classification unset on marking property

Ontology SDK

TypeScript Ontology SDK (OSDK) 2.0 is generally available the week of October 28th, offering a streamlined syntax to improve performance, enable more complex applications, and speed up the development process. By default, any new TypeScript applications created in Developer Console will use the latest 2.0.x generator. The [TypeScript OSDK 2.0 migration guide](/docs/foundry/ontology-sdk/typescript-osdk-migration/) explains version differences, highlights relevant syntax and structure changes, and provides code examples to help you update applications built using legacy versions. 




Introducing TypeScript Ontology SDK 2.0, now generally available

Quiver time series charts now have a more expansive set of next actions that allow you to create any time series-related entity directly from the next actions menu. Time series plots and event sets created from the next actions menu are added to the current chart by default, with the option to add to a new chart.

Expanded next actions menu for time series charts in Quiver

Improved object set aggregations by introducing a configurable limit for returned values. Users can now adjust the maximum number of results, which was previously fixed at 500. A new field has been added to the configuration to accommodate this change.



Enhanced object set aggregation limits

Pipeline Builder currently shows warnings in the deploy side bar when breaking changes will cause a full reprocessing of the upstream data. To further help users avoid accidentally reprocessing an entire pipeline, a new alert has been added to confirm that a deploy with reprocessing is intended.

Add replay confirmation alert for pipeline reprocessing

SAP connector

Version 2.32.0 (SP32) of the Foundry Connector 2.0 for SAP Applications add-on, used to connect Foundry to SAP systems, is now available for download from within the Palantir platform.

Foundry Connector 2.0 for SAP Applications v2.32.0 (SP32) is now available

Marketplace

DevOps

Marketplace now supports the [deletion of product installations](/docs/foundry/marketplace/installations/#deleting-installations). This feature allows users to delete deployed resources along with the installation location and resource.

Delete product installations

Workshop now supports the percentile, median, standard variation, and variance aggregations. These aggregations are only available for use on object types that use Object Storage V2. They can be used in [object set aggregation variables](/docs/foundry/workshop/concepts-variables/#variable-types), and also when configuring aggregations in the [XY Chart](/docs/foundry/workshop/widgets-chart/#chart-xy), [Vega Chart](/docs/foundry/workshop/widgets-vega-chart/#data-configuration-options), and [Pivot Table](/docs/foundry/workshop/widgets-pivot-table/#pivot-table) widgets.



New Aggregations in Workshop

Previously, users could not expand rows in pivot tables where the value was null. This has now been fixed, and users can examine the data in rows that are grouped on a null value.

Expandable rows in pivot tables support null values

When selecting write modes on Pipeline Builder outputs, an updated dialog now appears that describes the available modes with a helpful example image. Users can hover over other write modes to decide which one best fits their needs.


Updated write mode descriptions and examples

[Categories and tags](/docs/foundry/foundry-devops/manage-store-tags/) can now be used to group products together within a Marketplace store and filter against them when searching.


Categories and tags are now available

Functions

When authoring Python functions, Code Assist will be out of sync with the workstate less often, resulting in more accurate and timely error diagnostics and preview discovery.

Enhanced Code Assist and Preview reliability for Python functions

When authoring Python functions, issues with individual function signatures will be noted in code assist through clearer messaging. Other functions without issues can still be previewed as normal.

Enhanced preview for Python functions

Removed references to inherited API name for properties that are backed by shared properties within object types. You can no longer refer to properties via their shared property's API name within the context of an object type. You should instead use interfaces to enforce a common API name for a property across multiple object types.

Removed concept of inherited API name

Ontology Manager now ensures the **Allow multiple selections** setting is preserved when creating object types, and the property base type selection process is now more intuitive.

Property base type picker in object type creation wizard now more intuitive

Fix an issue where organizational submission criteria used to only work for users within a single Organization.

Organizational submission criteria fix

When creating a new value type, the base type will be prefilled based on context.  For example, if you have a property selected and want to create a new value type for it, the value type will automatically have the same underlying base property type as the property itself.

Set default type of new value type based on page

The Approvals inbox now retains user-set filters across sessions for a seamless experience. In the absence of saved filters, the inbox defaults to displaying access request task types.

Improved filter retention in the Approvals inbox

Sensitive Data Scanner

Users can now configure scans in Sensitive Data Scanner against Virtual Tables, extending the range of Sensitive Data Scanner to include federated data sources. Currently, Sensitive Data Scanner only supports scanning Virtual Tables through one-time scans. Users can configure one-time scans to include Virtual Tables by selecting it as a [transaction type](/docs/foundry/sensitive-data-scanner/create-a-sensitive-data-scan/#selected-transaction-types). [Review documentation on creating sensitive data scans](/docs/foundry/sensitive-data-scanner/create-a-sensitive-data-scan).

Sensitive Data Scanner can now scan Virtual Tables

Object Storage V2 now supports tracking user edit history. To use this feature, enable the **Track user edit history** setting in Ontology Manager  for object types, and update object views to include the **Edit History** widget to display user edit histories. The edit history reflects changes made to objects after enabling the **Track user edit history** setting. Any changes made prior to the activation of this feature will not be tracked. Additionally, user edit history from Object Storage V1 will not be available after migrating to Object Storage V2.

Support for edit history on Object Storage V2

Quiver now offers a **date difference** card that allows you to find the amount of time between two date parameters and specify a unit to express the difference in, such as days, weeks, or seconds.

Introducing the date difference card

Workshop now enables users to effortlessly record audio directly using their device's built-in microphone and seamlessly upload the recordings to a designated media set with the Audio Recorder widget.

Introducing the new Audio Recorder widget (beta)

Fixed an issue where useful error messaging was not surfaced if attempted resource imports failed to load into Pipeline Builder.


Error messaging for failed resource imports

When a property uses a value type, Ontology Manager now links directly to the value type from the property's **General** panel. This will open the value type in a new tab.


Direct links to value types in Ontology Manager

Users now have the ability to customize the number of relevant chunks returned during document retrieval augmented-generation in both [AIP Agent Studio](/docs/foundry/agent-studio/) and [AIP Threads](/docs/foundry/threads/). This feature allows users to adjust the amount of context provided to the model based on their specific workflow needs, potentially enhancing accuracy.


Configure the Number of Relevant Chunks Returned for Document RAG in AIP Agent Studio and AIP Threads

Artifact Repositories

Users can now personalize their Artifact repositories interface with a selection of themes:
- Light mode: Light mode can help with improved visibility in well-lit environments.
- Dark mode: Dark mode can help with reduced eye strain in low-light environments.
- System mode: System mode switches between light mode and dark mode based on system settings.

To choose a theme,  select the "Theme" dropdown menu in the upper left corner of a created Artifact repository. Then, choose a theme preference.

New themes for Artifact repositories

Due to its minimal usage and the significant impact on the performance and complexity of custom subgraphs, the AIP path prediction functionality has been discontinued in the Pipeline Builder.

Remove path predictions

Previously, pipeline creations would fail with a generic error message if the user did not have permission to create it in the provided directory. Now, the error messaging better communicates this failure to help users remediate the issue.

Improved error messaging for permissions issues

Marketplace products can now be exported from a local store and imported into another local store. Sharing is possible between any two local Marketplace stores on the same enrollment, or between two different enrollments. Review [documentation on the export and import of products](/docs/foundry/foundry-devops/export-import-products/) for an overview of the feature, including permission details.

Marketplace product exports and imports

The new Action type dialog now allows you to select your Organization membership as submission criteria for the Action type.

Add Organization submission criteria for new Action types

Previously, selecting the link to an AIP Logic-backed function within Ontology Manager redirected to Code Repositories. Now, this link correctly directs to the Logic function in the AIP Logic application.

Fixed AIP Logic redirect

Users can now create new value types from Ontology Manager using new buttons in the interface.


Create new value types

Users can now set up submission criteria that allow for all people in their organization to submit an action.

Allow organization Markings in action submission criteria

Control Panel has been revitalized with a sleek new look, featuring the organization and enrollment picker now conveniently located in the header for enhanced accessibility. The sidebar sports a fresh white color scheme, and first-time visitors to the Control Panel homepage will be greeted with an improved walkthrough experience, highlighting the revamped settings, favorites, and the new picker location.

Enhanced Control Panel accessibility and aesthetics

We made improvements to our error state communication to add detailed messaging, resolution prompts, and further debugging steps when needed. These errors states will now exist in the AIP Agent configuration window, conversation panel, save dialog, and message body, meaning users will no longer need to use browser debugging tools to solve Agent failures.

Improved error states

Workshop tooltips now feature the ability to incorporate line breaks, significantly improving the legibility of button descriptions. This enhancement ensures tooltips are more accessible and comprehensible for users.

Enhanced tooltip readability

Subtask statuses are now displayed as tags instead of icons, providing a more distinctive indicator of subtask progress.


Subtask statuses are now tags

We are excited to announce the beta release of custom-made LLM-powered assistants, also knows as AIP Assist Agents. This feature allows users to create LLM-powered assistants that specialize only on custom content sources, with no code or previous LLM experience required. This builds on the ability to register custom content sources with AIP Assist, which allows users to add Notepad documents or Rosetta custom documentation to AIP Assist's knowledge base. 

Create custom LLM-powered Assistants with AIP Assist Agents [Beta]

We added the ability to copy and paste components across different modules. You can copy a component or widget from one module, open a new Workshop module, then paste in your component.

Copy and paste across modules

New models are now available for use in Workshop LLM features, including  Llama 3.1 70B Instruct, Llama 3.1 80B Instruct, Anthropic Claude 3.5 Sonnet, Anthropic Claude 3 Haiku, GPT-4.o mini, and Mixtral 8x7B Instruct.


New LLM models available

Data Connection

Users can now create XML, HTML, and plain text bodies for webhook requests. 


New text body support webhook requests

Fixed an issue where schema checks would occasionally generate warnings that would not appear in the **Schema errors and warnings** dialog. Now, only warnings that appear in the dialog will be considered part of schema checks.

Consistent schema check warnings

A bug in Ontology Manager that prevented action tables from loading has been fixed. Previously, action tables did not load when a user did not have sufficient permissions to view an object type, or if an object type was deleted. Now, the action table will load and display the appropriate error. 

Bug fix for action tables in Ontology Manager

We are excited to introduce new modeling capabilities in Jupyter® Code Workspaces. Existing models can now be imported into Jupyter® notebooks, enabling a new series of interactive workflows in Code Workspaces, from fine-tuning existing models to integrating ML/AI capabilities into Code Workspaces dashboards.

Introducing model inputs in Code Workspaces

A refined creation and management process for code workspaces is now generally available. This new process allows Code Workspaces users to manage new workspaces from a single code repository resource, rather than from two distinct resources that need to be individually configured and permissioned. During code workspace creation, users now only need to configure the backing code repository, and new code workspaces can be accessed by selecting the code repository from anywhere in the Palantir platform filesystem, removing confusion and streamlining access. [Learn more about creating new code workspaces.](../code-workspaces/getting-started/)

Introducing a simplified process for creating and managing code workspaces

Users can now insert and remove rows from a table in Notepad.

Insert and remove table rows in Notepad

Users can can now add headers and footers when generating a Notepad template.

Add headers and footers in a Notepad template

Notepad now enables users to create multiple inline comments per paragraph to simplify collaboration, and the comment icon has been updated to improve visibility.

Multiple Inline Comments 🧵

Since July 2023, users have been able to leverage the power of AIP in Quiver to explore their data with ease by asking AIP to generate new or configure existing cards on their canvas via natural language prompts. Deploying the week of October 7th in beta on most AIP-enabled enrollments, Quiver's two primary large language model-driven capabilities - AIP Generate and AIP Configure - are now accessible beneath a card in addition to their existing availability in the top ribbon of Quiver's workspace and a selected card.




Improved access to and usability of AIP in Quiver [Beta]

Java support is generally available the week of October 7th in the [Ontology SDK](/docs/foundry/ontology-sdk/overview/) and [Developer Console](/docs/foundry/ontology-sdk/overview/#developer-console). Java developers can now generate a Java library to easily interact with Ontology objects, links, actions, aggregations, and functions. Users can add the [Java OSDK](/docs/foundry/ontology-sdk/java-osdk/) to existing Java services or create brand-new applications to enable seamless integrations with data on the Palantir platform.

Ontology SDK for Java Development [GA]

Notepad now features enhanced table management. Users can insert or remove multiple rows or columns using a new right-click menu as well as the existing actions toolbar. The number of rows or columns to modify can be changed by selecting the desired number of rows or columns on a table.

Improvements to table interactions in Notepad

Notepad users can now create multiple in-line comments and comment threads. Previously, users could only create one in-line comment thread, but it is now possible to create multiple threads by highlighting text and selecting the comment icon in the actions toolbar. Each comment bubble will display an individual comment thread when selected. Certain features, such as columns and tables, will only display a top-level comment button.

Notepad support for multiple in-line comment threads

Added a dynamic back button in the Ontology Manager that takes you back to the last significant page you visited. For example, if you navigate into a proposal, and then select an object type in the proposal, then select an action that edits that object type, using the back button will show you the object type. Once you use the button, it will take you to the proposal.  This is in contrast to using the browser's back button which will take you back to the last page you visited and not the parent in the hierarchy.

New and improved "Back" button

A bug affecting the ability to filter non-english merged proposals in Pipeline Builder has been fixed. All users now have the ability to filter merged proposals on the proposal page in all languages.

Bug fix for merged proposal filtering in Pipeline Builder

Users can now drag the output circle from one node to another to move all usages of an output. For example, if "dataset A" has three transforms that use it, you can drag the output circle on the graph from "dataset A" to "dataset B" so the transforms will use "dataset B". This behavior also works with transform outputs, which can be replaced by datasets or other transform outputs.

Allow output dragging to replace all usages

The Pipeline Builder interface now prominently displays tags to distinguish between importable and non-importable compute profiles, increasing user awareness. An acknowledgment step has been introduced for selecting non-importable profiles, ensuring deliberate user choices.

Improved compute profile permission visibility in Pipeline Builder

The Pipeline Builder  import compute profiles dialog has been improved. Previously, a list of profiles that the user did not have permissions to import were incorrectly displayed. This interface has been corrected and non-importable compute profiles are now adequately displayed. 

Improvements to the import compute profile interface in Pipeline Builder

Previously, when exiting the custom expression editor with unsaved changes, an alert would appear prompting users to save or discard changes. However, saving would not actually persist these changes. This issue has been fixed and changes will now persist before exiting.

Fixed bug where custom expression changes would not persist

Time Series

We are excited to announce the GA release of derived time series, available on all enrollments the week of October 7.  

With the GA release, you can now derive a new series from raw sensor data and save it back to the Ontology, allowing for the use of the derived series in all Ontology-aware applications. Once in the Ontology, derived series behave as raw time series but are calculated on the fly. This eliminates the need to manage or store derived data or duplicate those calculations across the platform.

Release notes

loading...