InRule

Onboarding Kit

What the platform is, how it works, and how to get your first rule running in 10 minutes. Load this into your LLM for context, or read it cover-to-cover before your first session.

What is InRule?

InRule is the browser-based tool for building, testing, and governing business rules. It is a decision intelligence platform used in banking, insurance, healthcare, government, and education finance, where decisions need to be defensible to regulators and auditors.

InRule sits between business intent and technical execution. Business analysts and technical users define decision logic using vocabulary and tables that read like natural language. When an application needs a decision, it calls a Decision Service and gets the answer back in milliseconds.

Named customers include HMRC, Bupa, Bank of America, and Allstate. InRule is used by over 500 companies worldwide.

How the platform works

Authoring

Where rules are created

InRule is the browser-based authoring environment. Business analysts, technical analysts, and developers build rule applications here. The AI assistant lives here too.

Catalog

Where they live and version

Every change creates a new version. Role-based access, full audit history, and promotion workflows are built in. The catalog must be hosted in InRule's SaaS environment to use InRule.

Execution

Where decisions run

Your calling application sends inputs to a Decision Service; the rules engine evaluates them and returns the decision. Each rule application gets a REST endpoint automatically on publish.

Core concepts

Rule application

The top-level project container. Holds the data model, rule sets, vocabulary, and test scenarios. Versioned as a unit so data and rules stay in sync.

Rule set

A named container for a group of related rules. The unit you author, manage, and promote between environments.

Rule types

Language rules: read like a sentence, for example "If all the following are true, set the value to X." The most common type, ~90% of most rule applications.
Decision tables: an Excel-style grid, conditions on the left, actions on the right. Exportable to and importable from Excel.
Syntax rules: more code-like, for complex formulas where operator precedence matters.

Entity

A grouping of related fields in the data model. A Customer entity might hold name, age, and address fields.

Field

A named, typed value on an entity. Types include text, number, decimal, date, yes/no, and value list.

Collection

A list of items. Used for inputs with multiple values (line items in a claim, borrowers on a loan) and for outputs (a list of decision reasons returned to the caller).

Vocabulary

Business-language constructs that let authors write rules in their own terms rather than technical field names. More vocabulary investment means more readable, maintainable rules.

Reference data

Predefined lookup data used by the rules. Lives inside the rule application (versioned with the rules) or in an external system accessed at runtime.

Getting started: a 4-chapter hands-on tour

2 min

Create a rule application

Start from blank, learn the structure, and know where rules live.

3 min

Build your first rule

Add a data model and a decision table inside the app you just created.

2 min

Test your rules

Run the rule against sample inputs and read the outcome.

2 min

Ask the AI assistant

Get a plain-language read on what your rule app actually does.

How your applications call decisions

Pattern	When to use it
REST API	The default. Every rule application in the catalog gets a REST endpoint automatically. Send a structured JSON payload, receive the decision back. Language-agnostic; works with any modern stack.
.NET SDK	For in-process execution inside your own .NET application, with no network round trip. Load the rule application into your process and call a method to get the result.
JavaScript	Rule applications export as JavaScript bundles. Run rules in the browser for instant form-level feedback, in serverless functions, or as Snowflake stored procedures.
Salesforce / Dynamics	First-class integrations with authoring extensions that bind tightly to those data models.

Deployment options

SaaS

InRule hosts the platform. Automated provisioning, high availability, zero infrastructure management. Required for InRule.

Self-hosted cloud

You run the platform in your own AWS, Azure, IBM Cloud, or Google Cloud account.

On-premises

You run it in your own data center. Full control over data and updates.

Hybrid

InRule SaaS for authoring and governance; your own runtime for data proximity and sovereignty.

AI assistant

InRule includes an AI assistant that helps rule authors work faster. It can find rules across a large rule application, summarize what they do, propose edits, and generate test data. Authors review and approve all AI suggestions. The AI never writes to the catalog without a human click.

The assistant is built on the AI Modeling Harness, which grounds AI behavior in InRule's actual data model and rule structure. It supports bring-your-own-LLM for customers who need to use their own approved AI provider.

Common questions

How long until my team can author rules independently?

Most authors complete their first production-style rule app in 4–6 hours of self-paced work, or a half-day live workshop. A 5-person team typically reaches independent authoring within two weeks of starting.

Do I need to know how to code?

No. InRule is built for business users. You build rules using vocabulary terms and decision tables that read like English. Technical staff handle integration, not authoring.

How do I test a rule before publishing?

Every rule app includes a test harness. Define sample inputs, run them through your rules, and inspect the outputs. Chapter 3 of the tour shows it end-to-end.

Where does my rule app live once I publish?

Published rule apps deploy to a Decision Service your applications call via API. You don't need to manage servers. Your Customer Outcome Manager (COM) helps set this up the first time.

Can I roll back a published rule?

Yes. Every publish creates a versioned snapshot. You can promote, demote, or roll back versions without redeploying anything downstream.

Who do I contact if I'm stuck?

Your Customer Outcome Manager (COM) is your first call. For platform-level issues, support@inrule.com is monitored 24/5.

How do I call a published rule app from my application?

Decision Services expose a REST endpoint per rule app. POST your input payload, receive the decision back as JSON. See the Decision Services API Guide for the full flow including authentication.

What authentication does the Decision Services API use?

OAuth 2.0 client credentials by default. Service principals are scoped per rule app. API keys are available for legacy integrations but not recommended for new work.

How do I version and promote rule apps across environments?

Rule apps version automatically on publish. Promotion between Dev, Stage, and Prod is a single API call or a click in the UI. CI/CD integrations exist for the major providers.

Is there an SDK for my language?

JavaScript and .NET are first-class. Java, Python, and Go can use the REST API directly. See the Decision Platform Developer Guide for typed clients and code samples.

Can I run InRule locally for development?

Yes. irAuthor Desktop authors against a local runtime, and the Decision Services Docker image runs anywhere. Most teams develop locally and deploy to managed Decision Services in production.

What's the difference between a rule set and a decision table?

A rule set is a named container for a group of related rules. A decision table is one type of rule inside a rule set: logic as a grid of conditions and actions. Most rule sets contain several rule types alongside decision tables.