Built with Rust

High-Performance Rules API

Deploy a headless evaluation service that reads rules published from BRMS. Built in Rust for maximum performance with automatic hot reloading when rules change.

How it works

Agent sits between your services and the rules storage, providing fast evaluation over HTTP.

BRMS
Storage
Agent
Your Services

BRMS publishes rules to storage. Agent watches for changes and hot-reloads automatically.

Built for Production

Enterprise-grade features for reliable rule evaluation at scale.

High Performance

Built in Rust for maximum throughput. Handle thousands of evaluations per second with minimal latency.

Hot Reloading

Rules update automatically when published from BRMS. No restarts required, zero downtime deployments.

Storage Agnostic

Reads rules from various storage backends. Works with S3, GCS, local filesystem, or custom providers.

Secure by Default

Sandboxed execution environment. Rules cannot access filesystem, network, or other system resources.

Observable

Built-in metrics and tracing. Monitor performance, track evaluations, and debug rule execution.

REST API

Simple HTTP API for rule evaluation. Easy to integrate with any service or application.

Use Cases

Common patterns for deploying Agent in your architecture.

Microservices Architecture

Deploy Agent alongside your services for centralized rule evaluation. All services call the same Agent instance, ensuring consistent decision-making across your system.

API Gateway Integration

Place Agent behind your API gateway for request validation, routing decisions, or rate limiting rules. Evaluate rules before requests reach your backend services.

Real-time Decisioning

Power real-time decisions like pricing, eligibility checks, or fraud detection. Sub-millisecond evaluation times enable use in latency-sensitive workflows.

Multi-tenant Applications

Serve different rules for different tenants from a single Agent instance. Rules are isolated and can be updated independently per tenant.

Ready to deploy?

Get started with Agent in minutes using Docker or Kubernetes.