Topics

The recurring questions, answered once and properly.

Long-form explanations of the concepts that come up in every AI operating decision — governed agents, connected context, execution continuity, operating systems versus point tools — written to be read rather than skimmed.

Concept-led

Organised by the question, not by our product

Durable

Written to stay correct past a release cycle

Linked

Each topic routes to the implementation path

Introduction

What belongs here and what does not.

A topic earns a page when the same question keeps arriving from people who are not asking about us. What is a business operating system. What makes an agent governed rather than merely permissioned. Why context is the constraint rather than model quality. Those deserve one careful answer that stays true.

Everything downstream of that — how a specific workflow is built, which connector to use, what a competitor does differently — lives in the implementation, integration, and comparison families, and each topic routes there rather than trying to be all of them.

The two product families, Launch and Grow, have their own topic hubs because the questions differ in kind: building an operating surface and running revenue against it are not variations on one subject.

Depth
Long-form, one concept per page
Scope
The concept, not the configuration
Product hubs
/topics/launch and /topics/grow

Why this exists

Why the concept layer keeps getting skipped.

AI content overwhelmingly answers "how do I do X with tool Y" and almost never answers "what is X, and is it the right thing to want". That is fine when the category is settled. It is expensive when the category is three years old and the vocabulary is still moving.

The cost shows up as decisions made on mismatched definitions: two people agreeing to build an agent while meaning different things by autonomy, or a platform bought on a definition of context the vendor did not share.

These pages exist so that the definitional argument happens before the procurement one, where it is cheap.

You're likely here because

  • Two teams are using the same word for materially different things
  • A decision keeps stalling on what counts as governed or autonomous
  • Vendor comparisons are being made on features rather than on the underlying model
  • The same explanatory conversation is happening in every meeting

How to choose

Where to start.

If

You are building the operating surface

If

You are running revenue and follow-up against it

If

You want the evidence rather than the explanation

If

You are comparing us against a specific product

How it works

How a topic page is built.

The same structure each time, so a topic can be entered at any point rather than read from the top.

01Define the term02State why it matters operationally03Show the system context04Give the honest limits05Route to the implementation

Step 01

Define the term

Plainly, and including what it excludes. A definition that covers everything decides nothing.

Step 02

State why it matters operationally

The decision the concept changes. If nothing downstream changes, it is vocabulary rather than a topic.

Step 03

Show the system context

Which systems participate, what stays authoritative, and where the boundary sits — the part that turns a concept into something implementable.

Step 04

Give the honest limits

What the approach does not solve. This is the section that makes the rest usable, and the one most often omitted.

Step 05

Route to the implementation

Every topic ends by pointing at the build, integration, or comparison page that acts on it, so the reading has somewhere to go.

All topics

Every topic, and the two product hubs — 59 in total.

Concept pages plus the Launch and Grow topic hubs, which carry the product-specific questions.

Product topic hubs

Building, and running revenue against what was built

Product topics

Building and running the operating surface.

Concepts

The vocabulary the decisions are made in.

Operating-system comparisons

Enterprise operating-system questions, answered for the mid-market.

What is Palantir?

Understand Palantir as an enterprise operating-system architecture for connecting data, AI, software delivery, and operational workflows.

Read the topic →

How Palantir works

Break down the operating model behind Palantir without reducing it to a dashboard or a single AI model.

Read the topic →

Palantir AIP explained

Understand AIP as the generative-AI layer in Palantir’s broader enterprise operating architecture.

Read the topic →

Palantir Foundry explained

Understand Foundry as the data and ontology foundation beneath many Palantir workflows.

Read the topic →

Palantir Ontology explained

Understand why Palantir’s Ontology is more than a semantic data catalog.

Read the topic →

Enterprise operating system

Understand the enterprise operating-system category that Palantir has helped make visible.

Read the topic →

AI operating system

Define what makes an AI operating system different from a chatbot, model endpoint, or dashboard.

Read the topic →

Palantir alternatives

Evaluate alternatives by the operating problem you need to solve rather than by copying Palantir feature for feature.

Read the topic →

Palantir alternatives for small businesses

Translate enterprise operating-system ideas into a smaller-business buying decision.

Read the topic →

Palantir for small business

Use Palantir’s operating-system concept as a framework for deciding what an SMB actually needs.

Read the topic →

Palantir for mid-market companies

Frame the operating-system decision for companies that have outgrown point tools but do not want unnecessary complexity.

Read the topic →

Palantir concepts for sales teams

Apply the connected operating-system model to pipeline, accounts, activity, forecasting, and next actions.

Read the topic →

Palantir concepts for marketing teams

Apply enterprise operating-system thinking to campaigns, content, attribution, audiences, and revenue handoffs.

Read the topic →

Palantir concepts for RevOps

Apply the enterprise operating-system pattern to the systems and handoffs that connect marketing, sales, and customer revenue.

Read the topic →

Palantir concepts for business operations

Apply the operating-system model to approvals, exceptions, handoffs, capacity, and recurring operational decisions.

Read the topic →

Palantir concepts for finance operations

Apply connected-data and workflow principles to planning, reporting, approvals, and operational finance without treating automation as financial advice.

Read the topic →

Palantir concepts for healthcare operations

Use the operating-system pattern for non-clinical healthcare workflows, capacity, scheduling, operations, and administration.

Read the topic →

Palantir concepts for pharmaceutical operations

Apply connected data, applications, and governed AI workflows to pharmaceutical business and operational processes.

Read the topic →

Palantir concepts for manufacturing

Apply ontology and operational AI concepts to production, quality, maintenance, inventory, and plant workflows.

Read the topic →

Palantir concepts for supply chain

Apply connected-data and decision workflows to inventory, suppliers, orders, logistics, and exceptions.

Read the topic →

Palantir concepts for logistics

Use operational AI patterns for routing, scheduling, capacity, service levels, and exception management.

Read the topic →

Palantir concepts for construction

Apply connected operational software to projects, schedules, crews, estimates, documents, and field handoffs.

Read the topic →

Palantir concepts for professional services

Translate the enterprise operating-system model into client delivery, knowledge, staffing, pipeline, and project workflows.

Read the topic →

Palantir concepts for real estate

Apply connected operations to leads, properties, transactions, scheduling, client communication, and portfolio workflows.

Read the topic →

Palantir concepts for insurance operations

Apply connected-data and workflow principles to intake, service, claims operations, sales, and internal coordination without automating regulated judgments by default.

Read the topic →

Palantir concepts for accounting firms

Apply a connected operating layer to client intake, document collection, status, reporting, and business-development workflows.

Read the topic →

Palantir concepts for agencies

Apply operating-system ideas to client delivery, campaign data, reporting, assets, pipeline, and recurring workflows.

Read the topic →

Palantir vs. AI app builders

Understand the difference between an enterprise operating system and a prompt-to-application builder.

Read the topic →

Palantir vs. no-code platforms

Compare no-code application creation with a broader enterprise operating-system model.

Read the topic →

Palantir vs. CRM platforms

Separate customer-system-of-record needs from broader enterprise operating-system needs.

Read the topic →

Palantir vs. business intelligence

Compare reporting and analytics with systems designed to connect decisions to operational actions.

Read the topic →

Dashboards vs. AI operating systems

Understand when visualization is enough and when the business needs a closed execution loop.

Read the topic →

AI ontology for business

Understand the role of a shared business-object model in operational AI.

Read the topic →

AI agents with enterprise data

Connect agents to governed business context without turning every data source into an unrestricted prompt feed.

Read the topic →

Operational AI explained

Define operational AI as AI connected to real workflows, systems, decisions, and actions.

Read the topic →

Closed-loop AI

Understand the difference between AI that produces an answer and AI that closes the business loop.

Read the topic →

Human-in-the-loop enterprise AI

Design AI workflows where people remain explicit decision makers at the points that require judgment, authorization, or accountability.

Read the topic →

Enterprise AI workflows

Build AI workflows around real systems, permissions, handoffs, and measurable outcomes.

Read the topic →

Build a Palantir-style operating system

Use the architectural ideas behind enterprise operating systems without assuming one vendor or recreating Palantir feature for feature.

Read the topic →

Palantir Pilot and AI app building

Understand Palantir’s move into natural-language application building on top of its Ontology.

Read the topic →

Scope

What is deliberately elsewhere.

  • Step-by-step implementation lives in the workflow, template, and build families, not here.
  • Connector specifics live in the integration guides, because they change on the provider's schedule rather than the concept's.
  • Head-to-head product judgements live in the comparison family, where the verdict can be stated plainly.
  • Measured findings live in research, which holds a stricter evidence standard than an explanatory page needs.

FAQ

Questions about this collection.

How is a topic different from a blog post?

A topic is maintained. It is expected to stay correct and to be revised when it stops being; a post is dated and stays as written.

Why are Launch and Grow separate hubs?

Because building an operating surface and running revenue against it raise different questions, and merging them produced pages that answered neither well.

Do these pages assume we use UbiGrowth?

No. The concept sections are written to be useful regardless, and the product appears where the implementation does — clearly marked as such.

How current are these?

They are revised when a topic stops being accurate rather than on a schedule. Where a topic depends on something that moves quickly, the page says so.

Start here

The definitional argument is cheaper than the procurement one.

Describe the decision you are trying to make. ARIA resolves which concepts actually bear on it and what the first bounded implementation would cover.