Deep Dive: Supercharge Your BI & Analytics

Achieving connected operations means going beyond the assembly line approach to data and analytics. It means creating full-fidelity feedback loops between your data, analytics, and operational workflows. Palantir Foundry can help make this possible, first by enriching your existing BI and analytics tooling and then by extending it to include more dynamic decision-making workflows.

Step 1: Enrich Your Existing BI Tooling with Foundry Data Integration and Ontology

Foundry supercharges your existing BI and analytic tooling with best-in-class data integration and ontology capabilities.

Foundry can integrate any existing data assets and models — from massive-scale sensor and IoT sources, telemetry and video data, licensed third-party data, and all internal sources. The Foundry Ontology sits on top of these integrated digital artifacts, using them to formulate a “digital twin” of the real-world things that constitute your business (e.g., plants, customers, transactions). The Ontology acts as a common data layer with multi-modal storage, supporting a diverse range of end-user functionality from search to time series analysis to geospatial visualization and more.

Together, these capabilities fortify conventional BI/analytics in several ways:

• Improve legibility and trust in data. Foundry offers interactive visibility into your data and model pipelines — allowing you to inspect the history of how a particular data set or piece of logic was generated. The platform automatically traces lineage from raw data all the way to applications where the data is used, eliminating the need to write pipeline lineage documentation that becomes obsolete when things change. As teams make decisions and generate new data on the operational front lines, the Ontology mediates interactions across layers of the stack and ensures that your user application layer can scale without fracturing your core data and model foundation.
• Incorporate robust security and governance capabilities. While many BI and analytics tools lack integrated security, Foundry offers unified security, lineage, and governance models from source to application. Every data artifact in Foundry can have a permissions and retention policy associated with it. Data administrators can apply governance policies to individual data sets, rows within data sets, folders, projects, and more — reducing the need to fork pipelines and providing a scalable, secure foundation for your BI applications.
• Embed read-write capabilities into your workflows. The Ontology orchestrates the flow of data, models, and actions across heterogeneous tooling and systems, allowing for continuous learning (discussed more below).

Foundry is designed to integrate deeply with your existing BI and analytical tools. It accomplishes this through REST APIs; ODBC and JDBC drivers; Python and R SDKs for connection to data science tools; and out-of-the-box connectors for common tools like Power BI and Tableau.

Step 2: Extend Your BI Ecosystem with Interactive Decision-making Applications

A fast-changing world demands the ability to pivot from “reactive” to “proactive” decision-making. Technologically, this means giving both technical and non-technical users not just ways to consume insights, but ways to manipulate, analyze, and act on heterogeneous data in dynamic, highly interactive applications.

As a result, we designed Foundry with capabilities that expand the pool of people who can transform data, perform analysis, build models, and build applications — reducing the bottleneck on engineering resources.

• Quiver is an analytical suite that allows you to plot, transform, and analyze massive-scale time series data and objects. Quiver is backed by time series infrastructure that balances high performance with a sustainable architecture that will scale as data volume and user head count grows, making Foundry both an effective historian and a real-time analytical tool for time series data. Quiver also provides object analysis capabilities to enable users to flexibly create and compare cohorts, perform aggregate analyses across object populations, and analyze object links and relationships.
• Contour is a point-and-click analysis, transformation, and reporting tool that excels at empowering non-technical users who have ad-hoc top down workflows. Contour allows you to quickly access data sets, conduct common analytical and logical operations in sequence to explore your data, debug data quality, and cleanse and transform your data. All of this can be done at scale — for instance, it allows analysis of large data sets with hundreds of millions of rows.
• Foundry Reports, Object Explorer, Fusion, and others give users endless options to further analyze and act on integrated data.

Beyond these out-of-the-box capabilities, Foundry recognizes that many valuable workflows are highly specific to your organization. To address this, we designed Workshop, a modular application builder that empowers a diverse community of builders to develop and manage complex read-write applications.

Workshop lets your teams compose applications using high-quality layouts and dynamic widgets, where builders articulate the inputs, outputs, and actions relevant to their use case. Here are just a few examples of interactive applications configured with Foundry’s modular application building blocks:

• Manufacturing Bill of Materials Application enables an organization’s raw material purchasing teams to run computations on spot-buy opportunities and calculate how cheaper materials could be utilized within the value chain — all while adjusting parameters like formulation constraints, existing inventory, and forecasted demand.
• Retail Common Operating Picture allows an organization’s store managers to overlay COVID case information atop retail store sales and operations data — and then adjust inventory and other plans as local case counts change.
• Airline Flight Alert Inbox shows aircraft swap managers a clear set of alerts on potential disruptions that require triage, review, and action. The inbox includes a detailed view of each individual alert, and action pathways allow users to make decisions on aircraft swaps and trigger changes in operational systems directly from the application.

Often, these applications are deployed in hours — fueling rapid concept iteration and saving weeks or months as compared to a fully bespoke solution. By abstracting away technical hurdles such as hosting, authentication, authorization, data integration, monitoring, and more, Workshop greatly reduces development costs and lets you to focus on building.

Step 3: Drive Org-Wide Decision-Making Processes

While the quality of each BI/analytic application matters, the connectivity of the system as a whole is equally important. Without a smart latticework across tools, organizations may see splintering in their data, logic, and processes — and in turn slowing progress in their ability to adapt to new conditions and improve.

To deliver deep operational connectivity across your BI, analytics, and operational tooling, Foundry employs an approach called decision orchestration. While conventional BI and analytics lack dynamic decision-based workflows, Foundry’s decision orchestration layer captures decisions made in user-facing applications and writes the results back into the organization’s Ontology, creating the feedback loops needed for high-velocity organizational learning.

Importantly, we designed Foundry so that it does not simply record past decisions, but actually enables you to simulate possible future states of the world. By leveraging Ontology-backed applications, decision makers can perform “what-if” analysis and scenario-test the downstream consequences of actions to better determine how they contribute to global business priorities.

1. Foundry digitally represents the chain of logistical, physical, and logical flows across organizational units and models how operations are functioning in near real-time.
2. Users can perform “what-if” analysis and simulate the downstream consequences of possible actions to better determine how decisions contribute to global business priorities.
3. When operators make decisions, the platform records the outcomes: which decisions were right, how they could have been better, and why they were made. This feedback loop allows the platform to continuously improve decisions and recommendations over time in alignment with global strategy.

To run these simulations, the Foundry Ontology blends inputs from integrated data and models, systems of action, and more. The platform then captures any decisions, results, and relevant context back into the Ontology as well as to underlying systems of action (e.g., ERP, CRM, MES, Asset Config, Edge). This operational feedback informs future simulations, improves data and AI/ML models, fuels higher-level decision-making, and accelerates learning within the organization.

See What Foundry Can Do For Your BI & Analytics Processes

Learn more about the Foundry BI & Analytics suite here, or explore Foundry's integration with popular SQL and BI tools here. Looking to start a conversation? Get in touch with a member of our Foundry product team to learn more about how Foundry can enhance your BI and Analytics tooling.