Application reference（应用程序参考）¶
You can interact with the Palantir platform using applications accessible via the sidebar. This page provides a reference for the range of applications available and describes when you may want to use each one.
Data connectivity and integration¶
Data connectivity and integration in the Palantir platform goes beyond traditional ETL/ELT solutions by offering a comprehensive suite of tools designed to enhance data team capabilities and reduce integration costs over time. Foundry serves as a robust data integration backbone for complex environments, providing an extensible framework that connects to various source systems and supports diverse data transfer methods. With integrated data transformation, management, and pipeline capabilities, Foundry ensures secure, scalable, and high-quality data delivery for critical operations.
| Application | Description | Use |
|---|---|---|
| Data Lineage | Data Lineage shows a graph of how different resources interacts with and flows through the platform. | Explore the origins or downstream usage of any data or resource in the Palantir platform. |
| Pipeline Builder | Pipeline Builder creates end-to-end pipelines from data sources to final outputs using LLMs and other built-in data transformations. | Integrate data towards analysis and application building with batch and streaming pipelines. |
| Code Repositories [1] [2] | Code Repositories is a web-based code authoring environment with support for versioning and collaboration. | Create data pipelines or write Functions in the Ontology. |
| Dataset Preview | Dataset Preview shows the contents and history of a dataset. | Browse a dataset and understand its history and other metadata. |
| Data Health | Data Health lets you define health checks to ensure datasets are high-quality. | Add or monitor health checks on datasets. |
| Data Connection | Data Connection allows you to connect to data sources and sync data into the Palantir platform. | Connect to organizational data sources or sync new datasets into the Palantir platform. |
| HyperAuto (SDDI) | HyperAuto generates end-to-end data pipelines on top of common ERP systems. | Generate an Ontology from enterprise systems without needing to develop pipelines manually. |
| Linter | Linter analyzes the state of your enrollment to identify anti-patterns and offers recommendations for optimizing resources, enhancing cost-efficiency, and improving pipeline stability and resilience. | Identify the highest impact recommendations, which can then be assigned to users for action using guided or automated workflows. Periodically review the expected impact using the impact summary feature. |
[1] Code Workbook or Code Workspaces may be more suitable for certain data science workflows. Learn more about the difference between Code Workbook, Code Workspaces, and Code Repositories.
[2] Pipeline Builder may be a better fit if you are a less-technical user.
Model connectivity and development¶
Model connectivity and development involve creating and integrating machine learning models that can be utilized across data pipelines, Ontology, and application layers to support diverse use cases. Models can be developed within the platform using open-source tools or imported from external environments, and once integrated, they benefit from platform features like inference, deployment, governance, and ML Ops. The platform ensures that every step of model development and deployment adheres to stringent standards for lineage, security, versioning, and auditing, enabling robust operationalization and continuous feedback loops for model improvement.
| Application | Description | Use |
|---|---|---|
| Model Assets | Model Assets enable integration of many model types into the Palantir platform. | Train models, and connect to externally hosted models in the Palantir platform. |
| Modeling Objectives | A modeling objective allows organizational stakeholders and model developers to collaborate on and deploy machine learning models. | Submit models; discuss modeling objectives, and deploy models into production. |
Ontology building¶
Ontology building in the Palantir platform involves creating an operational layer that maps datasets and models to their real-world counterparts, serving as a digital twin of the organization with both semantic and kinetic elements. By defining object types, link types, action types, and functions, the Ontology allows for robust end-user workflows, comprehensive metadata management, and dynamic security and governance. This integration empowers organizations to enhance decision-making at scale through Palantir's analytical and operational tools, enabling efficient data exploration, analysis, and application development.
| Application | Description | Use |
|---|---|---|
| Ontology Manager | Ontology Manager enables you to define your organization's Ontology. | Create new ontology resources such as objects, links or action types. |
| Object Views | Object Views represent the canonical way to display an object type. | Define user interfaces that can be used across use cases. |
| Object Explorer | Object Explorer allows you to visualize your Ontology. | Search, explore, and analyze your ontology. |
| Vertex | Vertex enables you to explore object relationships and run simulations. | Create system graphs of related objects and run end-to-end simulations using models. |
| Machinery | Machinery is a framework that enables the understanding and management of processes by identifying unwanted behaviors and facilitating improvements, while leveraging AIP's LLM capabilities to orchestrate multi-step automations in AIP workflows. | Use to optimize processes across various domains, such as procurement, sales, and customer service, by resolving inefficiencies, managing AI-driven workflows, mining event logs, monitoring performance metrics, and enabling real-time human intervention. |
| Foundry Rules | Foundry Rules enables users to actively manage complex business logic in the platform. | Create and apply rules to datasets, objects, and time series for a variety of use cases. |
| Map | Map provides powerful geospatial and temporal analysis and visualization capabilities. | Integrate data from across the platform into a cohesive geospatial experience. |
| Dynamic scheduling | Dynamic scheduling capabilities allow builders to create tailored scheduling and resource allocation workflows that accommodate the complex needs of their organization by leveraging the Ontology and modeling resource constraints and interdependencies. | Optimize employee schedules, enhance resiliency in transportation and logistics, and improve manufacturing output by providing interactive scheduling tools, real-time resolutions, and machine learning-backed recommendations. |
Developer toolchain¶
The Palantir platform's developer toolchain provides a comprehensive set of tools, including Core APIs and SDKs, development environments, and Compute Modules, enabling developers to build applications that leverage the Ontology's full capabilities. The Ontology SDK (OSDK) supports Python, Java, and TypeScript, allowing developers to interact with Ontology data and execute AIP Logic functions, while APIs facilitate programmatic management of platform access and data. Additionally, integrated development environments like VS Code and Compute Modules allow for scalable, containerized deployments, enhancing the development and integration of third-party applications and workflows.
| Application | Description | Use |
|---|---|---|
| Ontology SDK | The Ontology Software Development Kit (SDK) provides developers with direct access to the full capabilities of the Ontology from their development environment, supporting multiple languages and enabling seamless integration with Palantir APIs for application creation and management. | Accelerate application development by providing easy access to Ontology APIs, ensuring type-safety, reducing maintenance, and enhancing security, allowing developers to efficiently build applications powered by the Palantir platform. |
| Compute modules [Beta] | Compute modules allow you to deploy interactive containers on the Palantir platform, enabling you to run your existing code base, regardless of language, as serverless Docker images that can scale horizontally based on load in applications like Workshop and Slate. | Use compute modules to interact with code or third-party code by enabling container-backed functions, data integration, and hosting models, while integrating existing code bases in any language with dynamic scaling and connectivity. |
| Code Workspaces [3] | Code Workspaces brings the JupyterLab® and RStudio® Workbench third-party IDEs to Palantir. | Accelerate data science, statistics, and code-based workflows with industry-standard tools integrated into the Palantir platform. |
| VS Code workspaces | VS Code Workspaces integration on the Palantir platform uses the open-source VS Code to provide an IDE for writing and collaborating on production-ready code, with native support for Python transforms and OSDK React applications. | Efficiently develop code by accessing a pre-configured VS Code environment directly from within the Palantir platform, leveraging features like preview integration, debugging support, and automatic setup of Python environments, and get access to a pre-configured AI Coding Assistant which understands Foundry and its APIs. |
| Palantir extension for Visual Studio Code | The Palantir extension for Visual Studio Code integrates features from Code Repositories into VS Code, with a current focus on supporting Python transforms, providing a seamless development experience within the Palantir ecosystem. | Enhance the coding workflow by accessing Palantir repositories directly in VS Code, enabling features like authoring Python transforms and running Builds in Foundry. |
| Palantir MCP | Palantir MCP enables external AI IDEs and agents to connect to the Palantir platform and gain context on your Ontology and Foundry tools. | Use Palantir MCP to let external AI systems query data, access documentation, and build applications more efficiently. |
| Ontology MCP (OMCP) | Ontology MCP (OMCP) exposes application ontology resources as MCP tools, enabling external AI agents to read objects, execute actions, and query ontology data through controlled access. | Use Ontology MCP to enable external AI agents to safely interact with production ontology data through predefined actions and application restrictions. |
Use case development¶
The Palantir platform facilitates use case development by providing a suite of tools, such as application and workflow building tools, integrated analytics, and developer tools, all built on Foundry's core security and data management features. These tools allow teams to concentrate on enhancing operational capabilities, rather than dealing with infrastructure complexities, by continuously enriching a consistent set of data and model assets within the Ontology. This approach ensures that knowledge compounds as operational workflows are expanded across the enterprise, fostering scalable and efficient development.
Application building¶
| Application | Description | Use |
|---|---|---|
| Workshop [1] | Workshop enables the creation of interactive and high-quality applications for end users. | A point-and-click application builder that support pro-code customizations. |
| Slate [2] | Slate is an extensible application development framework. | Create a customized application using HTML, CSS, and JavaScript. |
| OSDK React applications | OSDK React applications allow developers to create fully customizable user interfaces using React, powered by the Ontology Software Development Kit, enabling seamless integration with Foundry as the backend for high-scale queries and secure application development. | Rapidly build and customize user interfaces by leveraging React's developer-friendly environment and Palantir's robust backend capabilities, ensuring efficient and secure application development in an organization. |
| Pilot | Pilot is an AI-powered application builder that creates applications from natural language prompts, automatically generating ontology entities, design specifications, frontend code, and seed data. | Describe the application you want to build in natural language and deploy production-grade hosted applications using OSDK and React through Developer Console. |
[1] Slate may be a better fit if heavy customization is needed for your application.
[2] Workshop is a better fit for applications of low to moderate complexity, and generally poses a lower maintenance cost over time.
Workflow building¶
| Application | Description | Use |
|---|---|---|
| Automate | Automate allows end users and application builders to see when data changes in the Palantir Ontology. | Configure automations to send notifications or submit Actions when certain conditions are met. |
| Solution Designer | Solution Designer is an interactive tool for crafting architectural representations of solutions on the Palantir platform, offering integration points, links to resources, and access to documentation and best practices. | Develop and refine solution architectures by exploring industry patterns, creating and comparing proposals, and facilitating project collaboration and knowledge transfer within an organization. |
| Carbon | Carbon lets you combine apps and other resources in the platform to create curated workspaces for end users. | Deliver a use case to end users that combines multiple applications or dashboards. |
Analytics¶
Palantir offers comprehensive analytical capabilities for all organizational users, integrating seamlessly with the Foundry Ontology to support both point-and-click and code-based analyses, such as table-based, geospatial, and temporal analysis. Its core applications—Contour, Quiver, Code Workbook, Notepad, and Fusion—facilitate diverse analytical tasks, from top-down data exploration and multimodal charting to machine learning and spreadsheet-driven computations, while enabling data to be written back into the Ontology for enriched insights. Additionally, Foundry supports deep integration with existing analytical tools through out-of-the-box connectors, REST APIs, ODBC/JDBC drivers, and SDKs for Python and R, enhancing connectivity with platforms like Power BI, Tableau, and Excel.
Learn more about analytical applications and the available types of analysis in the platform.
| Application | Description | Use |
|---|---|---|
| Contour [1] | Contour enables high-scale, top-down analysis on datasets. | To analyze tabular data in a point-and-click fashion. |
| Quiver [2] | Quiver enables analysis on object data and time series. | Analyze Ontology data and time series in a point-and-click fashion. |
| Insight [3] | Insight enables point-and-click analysis on Ontology objects with step-by-step analysis paths. | Perform analysis on Ontology data with link traversal, aggregations, visualizations, maps, SQL queries, and data writeback. |
| Code Workbook [Legacy] [4] | Code Workbook is a web-based environment for code-based analysis. | Analyze datasets in code, conduct data science workflows, or develop models. |
| Notepad | Notepad is an ontology-aware collaborative rich-text editor and document templating system. | Create rich-text documents with data and visualizations from Foundry applications. Use templates to add automated reporting to your workflows. |
| Fusion | Fusion is a bidirectional spreadsheet application for the Palantir platform. | Sync data from an editable spreadsheet to a dataset, or query and display data from the ontology in a spreadsheet. |
[1] Quiver or Insight may be a better fit for some workflows. Learn more.
[2] Contour may be a better fit for some workflows. Learn more.
[3] Insight is suited for analysis on known Ontology data with an intuitive workflow; Object Explorer is better for discovery and searching across unfamiliar Ontology data.
[4] Code Repositories and Pipeline Builder are recommended for developing production data pipelines. Learn more about Pipeline Builder and the difference between Code Workbook, Code Workspaces, and Code Repositories.
Product delivery¶
Foundry DevOps accelerates the development and deployment of data-driven workflows by packaging resources—such as ontologies, AI models, and pipelines—into flexible products, with automated version and dependency management ensuring seamless scalability. Key features include tagging product versions with release channels, managing a fleet of installations, and supporting use cases like product distribution, ecosystem building, release management, and bootstrapping new applications.
Marketplace further enhances product delivery by offering a storefront for easy discoverability and installation of published data products, complete with guided installation and related product recommendations.
| Application | Description | Use |
|---|---|---|
| DevOps | Create Foundry products for release management or workflow distribution. | Create Foundry products, manage installations, and release versions via release channels. |
| Marketplace | Discover and install Foundry products. | Customize installations of Foundry products, Examples, and workflow starter packs. Manage installations and configure upgrades to ensure installations stay up-to-date with minimal effort. |
Security and governance¶
Palantir’s platform is designed with security and governance as foundational pillars. This approach provides robust protection for sensitive data across industries such as healthcare, finance, and government. Moreover, Palantir's solutions are aligned with global regulatory frameworks like HIPAA, GDPR, and ITAR, ensuring comprehensive compliance and safeguarding of sensitive information.
The platform enforces strict access controls through a sophisticated security model that includes mandatory and discretionary permissions, granular access controls, and mandatory encryption, ensuring users only access authorized data. Palantir's enterprise security features include strong authentication measures, comprehensive audit logging, and extensive information governance.
| Application | Description | Use |
|---|---|---|
| Approvals | The Approvals application manages the workflow of requesting, approving, and invoking changes in Foundry, consolidating compliance, governance, and peer-review processes for efficient management of tasks like project access requests and ontology proposals. | Handle change requests in the Palantir platform by reviewing and approving tasks, tracking request status, and ensuring compliance with governance policies, all while staying informed through notifications about request progress and required actions. |
| Checkpoint | The Checkpoints tool is a data governance application that prompts users for justifications during potentially sensitive interactions, ensuring adherence to data protection, governance, and compliance policies through real-time review and auditability. | Customize prompts and gather justifications for user interactions, and review justifications in real-time and ensure compliance with organizational policies, while gaining insights into user behavior across various workflows. |
| Cipher | Cipher is a service that enables users to obfuscate data through cryptographic operations like encryption, decryption, and hashing. It provides easy-to-use tooling to manage algorithms and keys via Channels and Licenses, enabling the secure and accessible deployment of privacy-enhancing tools. | Apply cryptographic operations and manage permissions in workflows, enhancing privacy and governance by securely encrypting, decrypting, or hashing data within the Palantir platform's operational environment. |
| Sensitive Data Scanner | Sensitive Data Scanner (SDS) is a tool that enables organizations to discover and secure sensitive information within Foundry by specifying patterns of sensitive data to identify. It further allows organizations to automate appropriate actions to manage such data once detected, thereby reinforcing compliance with global data protection regulations. | Conduct singular or recurring scans across datasets to allow for enhanced control over sensitive data management and ensure compliance with privacy laws by automating the identification and handling of sensitive information. |
| Data Lifetime [Beta] | Data Lifetime is a service that enables the application of lineage-aware retention policies on datasets in the Palantir platform. This service governs the deletion of transactions and manages their downstream impacts, reinforcing compliance with the highest industry standards. | Set and manage retention policies that control the deletion of datasets and their downstream derivatives, ensuring comprehensive data removal and compliance with regulatory requirements by leveraging Foundry's data lineage capabilities. |
Management and enablement¶
Management and enablement capabilities are centralized within the Control Panel, offering a comprehensive suite for governance, resource management, and security administration, which supports enterprise data architectures like data mesh and data fabric. Control Panel allows administrators to configure and manage enrollments, map administrative functions to existing governance systems, and manage access through identity providers using SAML 2.0, ensuring seamless integration and security as the platform scales. Additionally, resource management tools provide granular insights into resource utilization, while customizable platform experiences ensure alignment with organizational branding and user-specific needs.
| Application | Description | Use |
|---|---|---|
| In-platform custom documentation | Custom documentation is a feature that allows users to create, publish, and manage documentation hosted on the Palantir platform, which can be indexed by AIP Assist to enhance query responses using your organization's documentation. | Develop and manage specific internal documentation within your organization, making it accessible through various platform entry points. |
| Walkthroughs | Walkthroughs is a tool that allows users to create custom, step-by-step tutorials on the Palantir platform, guiding end-users through applications or workflows with tailored, on-demand resources. | Design and share interactive tutorials that facilitate self-guided learning across multiple applications, incorporating rich media and progress tracking to enhance user engagement and support organizational workflows. |
Artificial Intelligence Platform (AIP) application reference¶
See the AIP application reference for a list of AIP applications.
中文翻译¶
应用程序参考¶
您可以通过侧边栏访问的应用程序(applications)与Palantir平台进行交互。本页面提供了可用应用程序范围的参考，并描述了您可能需要在何种场景下使用每个应用程序。
数据连接与集成¶
Palantir平台中的数据连接与集成超越了传统的ETL/ELT解决方案，提供了一套全面的工具，旨在增强数据团队能力并降低长期集成成本。Foundry作为复杂环境中的强大数据集成骨干，提供了一个可扩展的框架，能够连接到各种源系统并支持多种数据传输方法。凭借集成的数据转换、管理和管道功能，Foundry确保为关键操作提供安全、可扩展且高质量的数据交付。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| 数据血缘(Data Lineage) | 数据血缘以图形方式展示不同资源如何在平台中交互和流转。 | 探索Palantir平台中任何数据或资源的来源或下游使用情况。 |
| 管道构建器(Pipeline Builder) | 管道构建器使用大语言模型(LLMs)和其他内置数据转换功能，创建从数据源到最终输出的端到端管道。 | 通过批处理和流式管道，将数据集成到分析和应用程序构建中。 |
| 代码仓库(Code Repositories) [1] [2] | 代码仓库是一个基于Web的代码编写环境，支持版本控制和协作。 | 创建数据管道或在本体论(Ontology)中编写函数(Functions)。 |
| 数据集预览(Dataset Preview) | 数据集预览显示数据集的内容和历史记录。 | 浏览数据集并了解其历史记录和其他元数据。 |
| 数据健康(Data Health) | 数据健康允许您定义健康检查，以确保数据集具有高质量。 | 添加或监控数据集上的健康检查。 |
| 数据连接(Data Connection) | 数据连接允许您连接到数据源并将数据同步到Palantir平台。 | 连接到组织数据源或将新数据集同步到Palantir平台。 |
| HyperAuto (SDDI) | HyperAuto在常见ERP系统之上生成端到端数据管道。 | 从企业系统生成本体论(Ontology)，无需手动开发管道。 |
| 代码检查器(Linter) | 代码检查器分析您的注册(Enrollment)状态，以识别反模式，并提供优化资源、提高成本效益以及增强管道稳定性和弹性的建议。 | 识别影响最大的建议，然后可以通过引导式或自动化工作流将其分配给用户执行。使用影响摘要功能定期审查预期影响。 |
[1] 对于某些数据科学工作流，代码工作簿(Code Workbook)或代码工作空间(Code Workspaces)可能更合适。了解代码工作簿、代码工作空间和代码仓库之间的区别。
[2] 如果您是非技术用户，管道构建器(Pipeline Builder)可能是更好的选择。
模型连接与开发¶
模型连接与开发涉及创建和集成机器学习模型，这些模型可以在数据管道、本体论(Ontology)和应用层中使用，以支持多样化的用例。模型可以在平台内使用开源工具开发，或从外部环境导入，一旦集成，它们就能受益于推理、部署、治理和ML Ops等平台功能。该平台确保模型开发和部署的每一步都遵循血缘、安全、版本控制和审计的严格标准，从而实现强大的运营化和模型改进的持续反馈循环。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| 模型资产(Model Assets) | 模型资产支持将多种模型类型集成到Palantir平台。 | 训练模型，并连接到Palantir平台中外部托管的模型。 |
| 建模目标(Modeling Objectives) | 建模目标允许组织利益相关者和模型开发人员协作开发和部署机器学习模型。 | 提交模型；讨论建模目标，并将模型部署到生产环境。 |
本体论构建¶
Palantir平台中的本体论构建涉及创建一个操作层，将数据集和模型映射到其现实世界的对应物，作为组织的数字孪生，同时包含语义和动态元素。通过定义对象类型、链接类型、操作类型和函数(Functions)，本体论(Ontology)允许强大的最终用户工作流、全面的元数据管理以及动态的安全和治理。这种集成使组织能够通过Palantir的分析和操作工具增强大规模决策能力，实现高效的数据探索、分析和应用程序开发。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| 本体论管理器(Ontology Manager) | 本体论管理器使您能够定义组织的本体论(Ontology)。 | 创建新的本体论资源，例如对象、链接或操作类型。 |
| 对象视图(Object Views) | 对象视图代表了显示对象类型的规范方式。 | 定义可在不同用例中使用的用户界面。 |
| 对象浏览器(Object Explorer) | 对象浏览器允许您可视化您的本体论(Ontology)。 | 搜索、探索和分析您的本体论。 |
| Vertex | Vertex使您能够探索对象关系并运行模拟。 | 创建相关对象的系统图，并使用模型运行端到端模拟。 |
| Machinery | Machinery是一个框架，通过识别不良行为并促进改进来理解和管理流程，同时利用AIP的大语言模型(LLM)能力在AIP工作流中编排多步骤自动化。 | 用于优化采购、销售和客户服务等各个领域的流程，通过解决低效问题、管理AI驱动的工作流、挖掘事件日志、监控性能指标以及实现实时人工干预。 |
| Foundry规则(Foundry Rules) | Foundry规则使用户能够在平台中主动管理复杂的业务逻辑。 | 为各种用例创建规则并将其应用于数据集、对象和时间序列。 |
| 地图(Map) | 地图提供强大的地理空间和时间分析与可视化能力。 | 将来自平台各处的数据集成到一个连贯的地理空间体验中。 |
| 动态调度(Dynamic scheduling) | 动态调度功能允许构建者通过利用本体论(Ontology)并对资源约束和相互依赖关系进行建模，创建满足其组织复杂需求的定制化调度和资源分配工作流。 | 通过提供交互式调度工具、实时解决方案和机器学习支持的建议，优化员工排班、增强运输和物流的弹性，并提高制造产出。 |
开发者工具链¶
Palantir平台的开发者工具链提供了一套全面的工具，包括核心API和SDK、开发环境以及计算模块(Compute Modules)，使开发人员能够构建利用本体论(Ontology)全部功能的应用程序。本体论SDK (OSDK) 支持Python、Java和TypeScript，允许开发人员与本体论数据交互并执行AIP逻辑函数，而API则有助于对平台访问和数据进行编程管理。此外，VS Code等集成开发环境和计算模块(Compute Modules)允许进行可扩展的容器化部署，增强了第三方应用程序和工作流的开发与集成。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| 本体论SDK (Ontology SDK) | 本体论软件开发工具包(SDK)为开发人员提供了从其开发环境直接访问本体论(Ontology)全部功能的途径，支持多种语言，并能够与Palantir API无缝集成，用于应用程序创建和管理。 | 通过提供对本体论API的便捷访问，确保类型安全，减少维护工作并增强安全性，从而加速应用程序开发，使开发人员能够高效构建由Palantir平台驱动的应用程序。 |
| 计算模块(Compute modules) [Beta] | 计算模块允许您在Palantir平台上部署交互式容器，使您能够以无服务器Docker镜像的形式运行现有代码库（无论使用何种语言），这些镜像可以根据Workshop和Slate等应用程序中的负载进行水平扩展。 | 使用计算模块与代码或第三方代码交互，通过启用容器支持的函数、数据集成和托管模型，同时以动态扩展和连接性集成任何语言的现有代码库。 |
| 代码工作空间(Code Workspaces) [3] | 代码工作空间将JupyterLab®和RStudio® Workbench第三方IDE引入Palantir。 | 使用集成到Palantir平台中的行业标准工具，加速数据科学、统计和基于代码的工作流。 |
| VS Code工作空间(VS Code workspaces) | Palantir平台上的VS Code工作空间集成使用开源的VS Code，提供一个用于编写和协作生产级代码的IDE，原生支持Python转换(transforms)和OSDK React应用程序。 | 通过直接从Palantir平台内访问预配置的VS Code环境，利用预览集成、调试支持和Python环境自动设置等功能，高效开发代码，并访问一个了解Foundry及其API的预配置AI编码助手。 |
| Visual Studio Code的Palantir扩展 | Visual Studio Code的Palantir扩展将代码仓库(Code Repositories)的功能集成到VS Code中，当前侧重于支持Python转换(transforms)，在Palantir生态系统内提供无缝的开发体验。 | 通过在VS Code中直接访问Palantir仓库，增强编码工作流，实现编写Python转换和在Foundry中运行构建(Builds)等功能。 |
| Palantir MCP | Palantir MCP使外部AI IDE和代理能够连接到Palantir平台，并获取关于您的本体论(Ontology)和Foundry工具的上下文。 | 使用Palantir MCP让外部AI系统更高效地查询数据、访问文档和构建应用程序。 |
| 本体论MCP (Ontology MCP / OMCP) | 本体论MCP (OMCP) 将应用程序本体论资源暴露为MCP工具，使外部AI代理能够通过受控访问读取对象、执行操作和查询本体论数据。 | 使用本体论MCP使外部AI代理能够通过预定义的操作和应用程序限制，安全地与生产本体论数据进行交互。 |
用例开发¶
Palantir平台通过提供一套工具（如应用程序和工作流构建工具、集成分析工具和开发者工具）来促进用例开发，所有这些都建立在Foundry的核心安全性和数据管理功能之上。这些工具允许团队专注于增强运营能力，而不是处理基础设施的复杂性，通过在本体论(Ontology)中持续丰富一组一致的数据和模型资产来实现。这种方法确保了随着运营工作流在企业中扩展，知识得以累积，从而促进可扩展和高效的开发。
应用程序构建¶
| 应用程序 | 描述 | 用途 |
|---|---|---|
| Workshop [1] | Workshop能够为最终用户创建交互式、高质量的应用程序。 | 一个支持专业代码定制的点击式应用程序构建器。 |
| Slate [2] | Slate是一个可扩展的应用程序开发框架。 | 使用HTML、CSS和JavaScript创建定制化应用程序。 |
| OSDK React应用程序 | OSDK React应用程序允许开发人员使用React创建完全可定制的用户界面，由本体论软件开发工具包(Ontology SDK)驱动，能够与作为后端的Foundry无缝集成，支持高规模查询和安全应用程序开发。 | 通过利用React的开发者友好环境和Palantir强大的后端能力，快速构建和定制用户界面，确保组织内高效、安全的应用程序开发。 |
| Pilot | Pilot是一个由AI驱动的应用程序构建器，可以根据自然语言提示创建应用程序，自动生成本体论实体、设计规范、前端代码和种子数据。 | 用自然语言描述您想要构建的应用程序，并通过开发者控制台(Developer Console)使用OSDK和React部署生产级托管应用程序。 |
[1] 如果您的应用程序需要大量定制，Slate可能是更好的选择。
[2] Workshop更适合低到中等复杂度的应用程序，并且通常长期维护成本较低。
工作流构建¶
| 应用程序 | 描述 | 用途 |
|---|---|---|
| 自动化(Automate) | 自动化允许最终用户和应用程序构建者查看Palantir本体论(Ontology)中的数据何时发生变化。 | 配置自动化，以便在满足特定条件时发送通知或提交操作(Actions)。 |
| 解决方案设计器(Solution Designer) | 解决方案设计器是一个交互式工具，用于在Palantir平台上构建解决方案的架构表示，提供集成点、资源链接以及文档和最佳实践的访问。 | 通过探索行业模式、创建和比较提案，以及促进组织内的项目协作和知识转移，来开发和优化解决方案架构。 |
| Carbon | Carbon允许您组合平台中的应用程序和其他资源，为最终用户创建精心策划的工作空间。 | 向最终用户交付结合了多个应用程序或仪表盘的用例。 |
分析¶
Palantir为所有组织用户提供全面的分析能力，与Foundry本体论(Ontology)无缝集成，支持点击式和基于代码的分析，例如基于表格、地理空间和时间分析。其核心应用程序——Contour、Quiver、代码工作簿(Code Workbook)、Notepad和Fusion——促进了多样化的分析任务，从自上而下的数据探索和多模态图表绘制到机器学习和电子表格驱动的计算，同时支持将数据写回本体论以丰富洞察。此外，Foundry通过开箱即用的连接器、REST API、ODBC/JDBC驱动以及Python和R的SDK，支持与现有分析工具的深度集成，增强了与Power BI、Tableau和Excel等平台的连接性。
了解更多关于分析应用程序及平台中可用的分析类型。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| Contour [1] | Contour支持对数据集进行高规模、自上而下的分析。 | 以点击式方式分析表格数据。 |
| Quiver [2] | Quiver支持对对象数据和时间序列进行分析。 | 以点击式方式分析本体论(Ontology)数据和时间序列。 |
| Insight [3] | Insight支持通过逐步分析路径对本体论对象进行点击式分析。 | 对本体论数据进行分析，包括链接遍历、聚合、可视化、地图、SQL查询和数据写回。 |
| 代码工作簿(Code Workbook) [Legacy] [4] | 代码工作簿是一个基于Web的代码分析环境。 | 使用代码分析数据集、执行数据科学工作流或开发模型。 |
| Notepad | Notepad是一个感知本体论的协作式富文本编辑器和文档模板系统。 | 创建包含来自Foundry应用程序的数据和可视化的富文本文档。使用模板为工作流添加自动报告功能。 |
| Fusion | Fusion是Palantir平台的一个双向电子表格应用程序。 | 将数据从可编辑的电子表格同步到数据集，或从本体论查询数据并在电子表格中显示。 |
[1] 对于某些工作流，Quiver或Insight可能是更好的选择。了解更多。
[2] 对于某些工作流，Contour可能是更好的选择。了解更多。
[3] Insight适合对已知的本体论数据进行分析，具有直观的工作流；对象浏览器(Object Explorer)更适合在未知的本体论数据中进行发现和搜索。
[4] 建议使用代码仓库(Code Repositories)和管道构建器(Pipeline Builder)来开发生产数据管道。了解更多关于管道构建器以及代码工作簿、代码工作空间和代码仓库之间的区别。
产品交付¶
Foundry DevOps通过将资源（如本体论、AI模型和管道）打包成灵活的产品，加速数据驱动工作流的开发和部署，自动化的版本和依赖管理确保无缝的可扩展性。关键功能包括使用发布渠道标记产品版本、管理安装群组，以及支持产品分发、生态系统构建、发布管理和引导新应用程序等用例。
Marketplace通过提供一个商店界面进一步增强了产品交付，使已发布的数据产品易于发现和安装，并提供引导式安装和相关产品推荐。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| DevOps | 创建用于发布管理或工作流分发的Foundry产品。 | 创建Foundry产品，管理安装，并通过发布渠道发布版本。 |
| Marketplace | 发现并安装Foundry产品。 | 定制安装Foundry产品、示例和工作流入门包。管理安装并配置升级，以确保安装以最少的工作量保持最新状态。 |
安全与治理¶
Palantir平台的设计将安全和治理作为基础支柱。这种方法为医疗保健、金融和政府等行业的敏感数据提供了强大的保护。此外，Palantir的解决方案符合HIPAA、GDPR和ITAR等全球监管框架，确保全面合规并保护敏感信息。
该平台通过一个复杂的安全模型实施严格的访问控制，包括强制性和自主权限、细粒度访问控制和强制加密，确保用户只能访问授权数据。Palantir的企业安全功能包括强身份验证措施、全面的审计日志记录和广泛的信息治理。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| 审批(Approvals) | 审批应用程序管理在Foundry中请求、批准和调用变更的工作流，整合合规、治理和同行评审流程，以高效管理项目访问请求和本体论提案等任务。 | 通过审查和批准任务、跟踪请求状态以及确保遵守治理策略，来处理Palantir平台中的变更请求，同时通过关于请求进度和所需操作的通知保持信息同步。 |
| 检查点(Checkpoint) | 检查点工具是一个数据治理应用程序，在潜在敏感交互期间提示用户提供理由，通过实时审查和可审计性确保遵守数据保护、治理和合规政策。 | 自定义提示并收集用户交互的理由，实时审查理由并确保遵守组织政策，同时深入了解用户在各种工作流中的行为。 |
| Cipher | Cipher是一项服务，使用户能够通过加密、解密和哈希等密码学操作来混淆数据。它提供易于使用的工具，通过渠道(Channels)和许可证(Licenses)管理算法和密钥，从而安全且可访问地部署隐私增强工具。 | 在工作流中应用密码学操作和管理权限，通过在Palantir平台的操作环境中安全地加密、解密或哈希数据，增强隐私和治理。 |
| 敏感数据扫描器(Sensitive Data Scanner) | 敏感数据扫描器(SDS)是一种工具，使组织能够通过指定要识别的敏感数据模式，在Foundry中发现和保护敏感信息。它进一步允许组织在检测到此类数据后自动执行适当的管理操作，从而加强对全球数据保护法规的合规性。 | 对数据集进行单次或定期扫描，以增强对敏感数据管理的控制，并通过自动化识别和处理敏感信息来确保遵守隐私法律。 |
| 数据生命周期(Data Lifetime) [Beta] | 数据生命周期是一项服务，支持在Palantir平台的数据集上应用感知血缘的保留策略。该服务管理事务的删除及其下游影响，加强对最高行业标准的合规性。 | 设置和管理控制数据集及其下游衍生数据删除的保留策略，通过利用Foundry的数据血缘能力，确保全面的数据移除并符合监管要求。 |
管理与赋能¶
管理与赋能能力集中在控制面板(Control Panel)中，提供了一套全面的治理、资源管理和安全管理工具，支持数据网格和数据编织等企业数据架构。控制面板允许管理员配置和管理注册(Enrollments)，将管理功能映射到现有治理系统，并通过使用SAML 2.0的身份提供商管理访问，确保随着平台扩展的无缝集成和安全性。此外，资源管理工具提供对资源利用率的细粒度洞察，而可定制的平台体验确保与组织品牌和用户特定需求保持一致。
| 应用程序 | 描述 | 用途 |
|---|---|---|
| 平台内自定义文档 | 自定义文档是一项功能，允许用户创建、发布和管理托管在Palantir平台上的文档，这些文档可以被AIP Assist索引，以使用您组织的文档增强查询响应。 | 在组织内开发和维护特定的内部文档，使其可通过各种平台入口点访问。 |
| 引导教程(Walkthroughs) | 引导教程是一种工具，允许用户在Palantir平台上创建自定义的、逐步的教程，通过定制的、按需的资源引导最终用户使用应用程序或工作流。 | 设计和分享交互式教程，促进跨多个应用程序的自导学习，结合富媒体和进度跟踪，以增强用户参与度并支持组织工作流。 |
人工智能平台(AIP)应用程序参考¶
请参阅AIP应用程序参考以获取AIP应用程序列表。