AI-Powered Decision Platform

Code Effects 6 is the industry-first LLM-agnostic decision engine, designed to be compiled and deployed across a wide range of modern and legacy environments. By combining the performance of IL-compiled execution with the flexibility of AI-guided authoring using the LLM of your choice, it eliminates the “black box” of legacy systems through native Visual Studio debugging and replaces rigid decision tables with a context-aware, adaptive UI.

Hero

Why Code Effects?

  • First true BYOAI: Use any LLM of your choice in both rule authoring and evaluation. No restrictions. Seriously.
  • First native VS debugging: Pass a delegate to the Evaluate method and set breakpoints - just like in any code.
  • Unique rule authoring UI: Embeddable into any web app. Intuitive enough for end users with little or no training.
  • Perpetual licensing: Pay only once. No recurring fees. Upgrades are always optional. No more “SaaS tax.”
  • Data security: No external calls. Just reference two libraries from NuGet. Data and logic stay within your environment.
  • Runs anywhere: With a Full Source perpetual license, compile and deploy on virtually any platform and hardware.
  • Wide adoption: Used by major U.S. insurance and audit firms, global lenders, and government agencies. Join them.

AI Without Vendor Lock-In

The Adaptive Source feature connects the Rule Editor to any LLM - OpenAI, Claude, Gemini, or private models - through a client-defined callback that receives the full rule context as the user builds the rule. This enables precise control over UI context menus, and how AI influences rule authoring, without dependence on a specific LLM. It also allows AI behavior to be tailored to your domain, ensuring consistent and context-aware rule creation.

AI-Integrated Rule Evaluation

The engine combines deterministic rule execution with probabilistic AI-driven outputs, allowing structured decision logic to incorporate predictions, scores, percentages, and boolean results directly into evaluation flow. Prompt-enabled methods can invoke any LLM and return values that participate in standard rule processing. This creates a unified model where traditional business logic and AI inference operate together within the same execution pipeline.

Perpetual Ownership

The Platform uses a transparent, perpetual licensing model with no recurring fees. You pay once and retain full ownership of the software, with optional upgrades to future major versions. With no per-user, per-CPU, or per-evaluation charges, it scales predictably for enterprise applications. This model offers a clear cost advantage over typical SaaS solutions by eliminating subscription overhead and enabling predictable long-term budgeting.

Security-First Data Architecture

The engine operates as a self-contained library within your application, with no required external services or external dependencies. Rules, data, and user context remain within your environment, with any external integrations defined by your logic or implementation fully controlled by your team. This architecture supports secure, compliant deployments, including isolated or air-gapped systems, and aligns well with modern data protection requirements.

True Native Rule Debugging

Native rule debugging is available directly within Visual Studio by passing a tracing delegate to the Evaluate method. This allows you to set breakpoints, step through rule execution, inspect condition-level results alongside the source object in real time, or record detailed execution traces in production. The result is full transparency into rule behavior and reduced troubleshooting effort - the first native debugger and logger of its kind in business rules.

Endless Customization

Rule authoring and evaluation can be tailored to your application’s requirements. The Rule Editor supports dynamic multilingual translation of UI elements, including fields, methods, labels, and messages. Evaluation and authoring behavior can be customized through value ranges, menu filtering, grouping, rounding, string handling, and parentheses logic, among many other configuration options, to name just a few.

Native-Level Evaluation Speed

Rules are compiled directly into IL rather than interpreted at runtime, making them a native part of your application’s execution path. This eliminates the overhead of interpretation layers and enables high-performance, low-latency rule evaluation. Performance scales with your infrastructure, supporting the evaluation of large volumes of complex rules with predictable throughput, even under sustained peak loads.

Environment-Agnostic

Built on .NET Standard 2.0 specification, Code Effects 6 supports deployment across Windows, Linux, macOS, iOS, and Android on any compatible .NET runtime. With an available Full Source perpetual license, teams can compile and optimize our engine for their specific environments, from containerized Linux deployments with modern .NET versions to existing .NET Framework applications and long-term legacy systems.

Model-Agnostic AI Architecture

Prompt-enabled inputs allow your rules to work with any LLM, making AI a natural part of rule evaluation. These inputs can be passed into your in-rule methods or rule actions, where they can invoke a model of your choice and return values such as probabilities, percentages, or boolean results. This makes AI a flexible, pluggable component within your decision logic.

LLM Architecture

AI as Architecture, Not Integration

Code Effects 6 does not treat AI as a prebuilt model, plugin, or API hook. Instead, it introduces a contract-level abstraction that decouples the engine from any specific LLM, enables AI to be injected at the method level, and preserves the integrity of rule structure while making evaluation fully extensible.

AI Prompt

Prompt as a First-Class Rule Element

The Platform introduces a Prompt flag that can be applied to any string property, method return value, or parameter of your source object. This makes AI part of the rule grammar rather than an external service, allowing user-defined models to participate directly in both rule authoring and execution.

Deterministic Results

Deterministic + Probabilistic Decisions

The engine blends deterministic rule evaluation with probabilistic outputs generated by LLMs of your choice. Deterministic logic defines and orchestrates decision flow, while probabilistic results, such as predictions or scores, are incorporated directly into that flow as condition values.

Our Customers

  • Chevron Logo
  • Barclays Logo
  • The Home Depot Logo
  • Ernst and Young Logo
  • Equifax Logo
  • DuPont Logo
  • Reuters Logo
  • PWC Logo
  • Verizon Logo
  • MetLife Logo
  • Moodys Logo
  • Aon Logo
  • MassMutual Logo
  • New York Life Logo
  • Humana Logo
  • Chevron Logo
  • Barclays Logo
  • The Home Depot Logo
  • Ernst and Young Logo
  • Equifax Logo
  • DuPont Logo
  • Reuters Logo
  • PWC Logo
  • Verizon Logo
  • MetLife Logo
  • Moodys Logo
  • Aon Logo
  • MassMutual Logo
  • New York Life Logo
  • Humana Logo

Business

Whether you have a small web form that requires flexible data validation or you are building advanced AI-driven systems, Code Effects provides a unified approach to decision automation. Combining deterministic rule execution with AI-augmented logic, it adapts to both simple and complex scenarios across modern and legacy environments. You can explore these capabilities through our live demo, demo projects, or product documentation.

Integration

Built on .NET Standard 2.0, our rules engine integrates directly into your applications without complex setup or external dependencies. It consists of a web-based Rule Editor for managing rules and a high-performance Rules Engine for executing them, both of which can run together in the same project or be deployed independently across different systems and environments.

Philosophy

The Platform replaces traditional decision tables with a flexible, context-aware Rule Editor that enables business users to define and manage rules without relying on IT or specialized formats. Its intuitive interface allows rules to be expressed in familiar terms, improving usability. Combined with a thread-safe, high-performance engine, it supports large-scale rule evaluation with consistent, low-latency execution across a wide range of scenarios.

l102

p101

×