From Intent to Action with Three Cortex Routers

Coordinating models, search, and engineering expertise through one Enterprise AI experience

Enterprise AI is moving beyond the single-model assistant.

Software work spans coding, testing, security, research, review, remediation, and verification. Although these activities may begin in the same chat interface, they require different forms of intelligence and context.

Cortex addresses this through three complementary capabilities:

1. Model Routing connects requests with the appropriate Cortex intelligence.
2. Search Routing provides access to current public information across multiple search services.
3. Skill Routing brings focused engineering practices into relevant workflows.

Together, they help Cortex understand not only what a user is asking, but also what type of intelligence, information, and engineering expertise can best support the outcome.

The result is a simpler experience for users and a more adaptable AI foundation for organizations.

One experience, specialized intelligence

AI platforms are becoming more capable, but added capability can create added complexity.

Users may be asked to choose between models, agents, tools, search modes, and large collections of specialized instructions. That exposes the internal architecture of the AI system to the person simply trying to complete a task.

Cortex takes a different approach.

Developers and enterprise users express what they want in natural language. Cortex coordinates the appropriate capabilities behind a unified interface.

A request might involve:

• Understanding unfamiliar code
• Researching current technical information
• Implementing a feature
• Reviewing a proposed change
• Investigating a security concern
• Improving performance
• Creating tests
• Verifying whether a patch satisfies its objective

The experience remains consistent even though the intelligence supporting each request may differ.

This reflects a central principle of the Cortex AI Model Ensemble: specialized capabilities should improve the result without forcing users to manage the specialization themselves.

Model Routing

Cortex Router 1.7 established the coordinating entry point for the Cortex AI Model Ensemble. It helps Cortex recognize the nature of a request and connect it with the appropriate model, agent, or product experience. Introducing Cortex Router for the AI Model Ensemble

Model Routing allows Cortex to support different kinds of work through one interface.

A quick technical question may need a direct response. A larger engineering objective may benefit from coding assistance or structured planning. A security-sensitive request may require specialized analysis. A completed change may benefit from independent verification.

For customers, the value is straightforward:

• Less time choosing between models or operating modes
• More consistent handling of different engineering requests
• Better use of specialized Cortex intelligence
• A clearer path from a user’s objective to an appropriate workflow
• Greater separation between coding, security, safety, and verification responsibilities

Model Routing helps transform the Cortex Ensemble from a collection of specialized capabilities into one coordinated product experience.

The user focuses on the outcome. Cortex handles the complexity of selecting the appropriate intelligence.

Search Routing

Many software and business questions depend on current information.

Developers need recent framework documentation, API references, release notes, compatibility guidance, and technical examples. Security teams may need current vulnerability information or vendor advisories. Enterprise users may need public information that changes more quickly than a model’s built-in knowledge.

Cortex Search Routing expands public information access across four search services:

• Google
• Exa
• Brave Search
• DDGS

The objective is not to make customers choose a search engine for every question. It is to give Cortex a broader and more flexible search foundation.

Search Routing helps Cortex deliver:

• Access to current public information
• Stronger support for technical and documentation research
• Multiple search options within one Cortex experience
• Broader public-information access
• Search availability across different Cortex packages
• Clearer visibility into search usage

This capability separates the user’s research need from any single search provider.

A developer can ask Cortex to find current SDK documentation without leaving the engineering workflow. A security professional can research a recently disclosed issue while continuing the investigation in Cortex. A user can request current general information without manually opening separate search tools and transferring the results back into the conversation.

Search becomes part of the coordinated Cortex experience rather than a separate destination.

Current information in context

Retrieving information is only part of the problem.

The larger value comes from connecting that information to the user’s objective. Public documentation may need to inform an implementation. A release note may explain a regression. A security advisory may need to be considered alongside the organization’s code and configuration.

Search Routing allows current public information to participate in a broader Cortex workflow while preserving a consistent user experience.

The customer asks the question once. Cortex can combine relevant public information with the other capabilities involved in completing the task.

Skill Routing

Models provide broad intelligence. Skills provide focused engineering guidance.

Cortex Skill Routing helps bring relevant software-development practices into a request without requiring the user to manually locate and attach specialized instructions.

The capability is organized around major areas of secure software delivery:

• Testing
• Security Analysis
• Security Validation
• Implementation
• Verification

These areas cover practices such as test-driven development, threat modeling, security review, vulnerability analysis, secure remediation, refactoring, performance optimization, patch verification, code review, and regression analysis.

The customer does not need to navigate this internal organization.

A developer can simply ask:

• Use TDD to implement this function.
• Review this change for security weaknesses.
• Refactor this component without changing behavior.
• Verify that this patch addresses the reported issue.
• Add regression coverage for this bug.
• Analyze whether the change introduces unintended effects.

Cortex can recognize the relevant engineering discipline and apply appropriate guidance to the work.

Focused expertise when it matters

Skill Routing is designed to make Cortex more consistent, not more complicated.

For customers, it provides:

• Better alignment with established engineering practices
• More consistent testing and implementation workflows
• Stronger security-oriented development guidance
• Clearer separation between implementation and verification
• Reusable organizational expertise
• A foundation for future team-specific engineering practices

This is particularly valuable for organizations seeking consistency across teams.

Engineering standards often exist across documentation, review checklists, security requirements, and institutional knowledge. Skill-based assistance creates a path for relevant guidance to become part of the AI-assisted workflow rather than remaining disconnected from it.

Cortex can support focused practices while keeping the user’s interaction simple.

Three connected layers

Model Routing, Search Routing, and Skill Routing each solve a different customer problem.

Together, they create a more complete Enterprise AI experience.

Consider a developer asking Cortex to investigate a problem involving a recently updated software library.

Model Routing connects the request to the appropriate Cortex capability. Search Routing can bring current technical information into the investigation. Skill Routing can apply the relevant testing, implementation, security, or verification practice.

The developer does not need to coordinate these capabilities manually.

This is the larger Cortex direction: one interface at the surface, coordinated specialization underneath.

Built for enterprise use

Routing is not only about improving responses. It also creates a stronger foundation for enterprise control.

As organizations adopt more AI models, information services, and engineering workflows, they need a coherent way to bring those capabilities together. Otherwise, every new capability becomes another isolated integration that users and administrators must manage.

The Cortex routing architecture helps organizations:

• Use specialized AI without exposing unnecessary complexity
• Connect current information with engineering work
• Apply more consistent development and security practices
• Expand capabilities without redesigning the user experience
• Maintain clearer boundaries between different AI responsibilities
• Support a more adaptable Enterprise AI strategy

This matters because AI platforms will continue to evolve.

Models will change. Search services will expand. Organizations will develop new engineering standards. Security requirements will become more demanding. The underlying platform must be able to evolve without forcing users to continually relearn how to access its capabilities.

Routing provides that connective layer.

Part of the Cortex journey

These routing capabilities build on the broader evolution of Cortex.

The Cortex AI Model Ensemble introduced specialized intelligence across secure software development. Cortex Router created a coordinated entry point. Cortex Planner added structure for larger engineering objectives. Cortex Connect extended the experience across browser, mobile, and a connected Visual Studio Code workspace.

Model Routing, Search Routing, and Skill Routing deepen that foundation.

They help Cortex connect three essential elements:

• The appropriate intelligence
• The appropriate information
• The appropriate engineering expertise

That connection can follow the user across the Cortex experience.

A request may begin in the browser while reviewing documentation. It may continue on mobile as the user follows progress. It may reach a connected development workspace when project-aware work is required.

The work surface can change while the objective remains coherent.

Toward coordinated outcomes

The first generation of AI assistants focused on producing answers.

The next generation added code generation and tools.

Enterprise AI must go further. It must coordinate models, information, engineering practices, connected systems, and user control around meaningful outcomes.

Cortex is building toward that coordinated experience.

Model Routing helps connect intent with specialized intelligence.

Search Routing brings current public information into Cortex through a broader search foundation.

Skill Routing adds focused engineering practices when they are relevant.

The value is not the number of models, providers, or skills available behind Cortex. The value is how naturally they work together for the customer.

One request. One Cortex experience. The right intelligence, information, and expertise brought together around the outcome.