Skip to main content

Examples Overview

Complete, runnable examples for common use cases - Get started quickly with production-ready code.

Available Examples​

Browse our collection of production-ready examples. Each example is complete, tested, and ready to deploy.

Quick Reference​

ExampleIndustryComplexityFeaturesTime to Deploy
Hello WorldAllBeginnerBasic query + policy5 minutes
Customer SupportSupportIntermediatePII redaction, RBAC, audit10 minutes
Banking AI AssistantBankingAdvancedRBI, PCI-DSS, Fraud Detection, PII45 minutes
Healthcare AssistantHealthcareAdvancedHIPAA, PII, RBAC30 minutes
E-commerce RecommendationsRetailIntermediateProduct data, personalization20 minutes
Trip PlannerTravelAdvancedMAP, MCP, LLM integration30 minutes

Hello World​

The simplest AxonFlow example - Perfect for learning the basics.

What It Demonstrates​

  • Basic query execution
  • Simple policy enforcement
  • Request/response handling
  • Audit logging

Code​

TypeScript (30 lines):

import { AxonFlowClient } from '@axonflow/sdk';

const client = new AxonFlowClient({
  endpoint: 'https://YOUR_AGENT_ENDPOINT',
  licenseKey: 'YOUR_LICENSE_KEY',
  organizationId: 'my-org'
});

async function main() {
  const response = await client.executeQuery({
    query: 'What is the capital of France?',
    policy: `
      package axonflow.policy
      default allow = true
    `
  });

  console.log('Response:', response.result);
  console.log('Latency:', response.metadata.latency_ms + 'ms');
}

main();

Go (35 lines):

package main

import (
    "context"
    "fmt"
    "log"

    "github.com/getaxonflow/axonflow-sdk-go"
)

func main() {
    client, _ := axonflow.NewClient(axonflow.Config{
        Endpoint:       "https://YOUR_AGENT_ENDPOINT",
        LicenseKey:     "YOUR_LICENSE_KEY",
        OrganizationID: "my-org",
    })

    response, err := client.ExecuteQuery(context.Background(), &axonflow.QueryRequest{
        Query: "What is the capital of France?",
        Policy: `
            package axonflow.policy
            default allow = true
        `,
    })
    if err != nil {
        log.Fatal(err)
    }

    fmt.Println("Response:", response.Result)
    fmt.Printf("Latency: %dms\n", response.Metadata.LatencyMS)
}

Quick Start​

# Clone example
git clone https://github.com/axonflow/examples
cd examples/hello-world

# TypeScript
cd typescript
npm install
npm start

# Go
cd go
go run main.go

GitHub​

📁 View on GitHub

Banking AI Assistant​

RBI/PCI-DSS compliant banking AI assistant - Production-ready fraud detection and loan processing.

What It Demonstrates​

  • Real-time fraud detection with AI-powered pattern analysis
  • Loan application processing with instant preliminary assessment
  • PII protection (SSN, account numbers) automatically
  • Regulatory compliance (ECOA, FCRA, TILA, RBI)
  • Complete audit trails for compliance requirements
  • In-VPC deployment for data sovereignty

Features​

✅ Fraud Detection:

  • Impossible travel detection
  • Unusual transfer monitoring
  • Card cloning detection
  • Risk scoring (0-10 scale)

✅ Loan Processing:

  • Debt-to-income ratio analysis
  • Income multiple validation
  • Compliance checking (ECOA, FCRA, TILA)
  • Instant preliminary assessment

✅ PII Protection:

  • Automatic SSN redaction
  • Account number masking
  • Credit card protection
  • Contact information partial masking

✅ Compliance:

  • RBI guidelines built-in
  • PCI-DSS support
  • Complete audit logging
  • CloudWatch integration

Architecture​

Quick Start​

Note: Full implementation available to AWS Marketplace customers. See Banking Example for details.

# Clone repository (customers only)
git clone https://github.com/axonflow/examples
cd examples/banking-assistant

# Setup
npm install
docker-compose up -d

# Run
npm run dev

# Access at http://localhost:3000

Policy Example​

package axonflow.policy.banking

import future.keywords

# Block suspicious transactions
deny["Suspicious: impossible travel"] {
    input.transactions[i].location != input.transactions[j].location
    time_diff := abs(input.transactions[i].timestamp - input.transactions[j].timestamp)
    time_diff < 300  # 5 minutes
}

# Redact SSN from all queries
redacted_query := regex.replace(
    input.query,
    `\b\d{3}-\d{2}-\d{4}\b`,
    "***-**-****"
)

# RBI compliance - data sovereignty
deny["RBI violation: cross-border data transfer"] {
    input.context.data_location != "IN"
    input.context.customer_region == "IN"
}

Performance​

OperationLatencyNotes
Fraud pattern analysis<50msReal-time detection
Loan assessment<100msFull DTI calculation
PII detection<5msAutomatic redaction
Policy enforcement<10ms P95Single-digit ms typical

GitHub​

📁 View on GitHub

Documentation​

📖 Full Documentation

Healthcare Assistant​

HIPAA-compliant medical AI assistant - Production-ready healthcare example.

What It Demonstrates​

  • HIPAA compliance patterns
  • PII detection and redaction
  • Role-based access control (RBAC)
  • Patient data access controls
  • Multi-agent coordination
  • Audit trail for compliance
  • HL7/FHIR integration patterns

Features​

✅ HIPAA Compliance:

  • Encryption at rest and in transit
  • Access controls and audit logging
  • Minimum necessary rule enforcement
  • PHI (Protected Health Information) protection
  • Breach notification procedures

✅ Security:

  • Role-based permissions (Doctor, Nurse, Admin)
  • Patient assignment validation
  • Business hours enforcement
  • Emergency access procedures
  • Comprehensive audit trail

✅ Technical:

  • React frontend with TypeScript
  • Go backend with PostgreSQL
  • AxonFlow policy enforcement
  • Real-time audit logging
  • CloudWatch integration

Use Cases​

  1. Patient Records Access

    • Doctors access assigned patients only
    • Nurses have limited read access
    • Admins have full access with audit trail

  2. Prescription Management

    • Drug interaction checking
    • Dosage validation
    • Prescription history tracking

  3. Lab Results

    • Automated result distribution
    • Critical result alerts
    • Access based on patient assignment

  4. Appointment Scheduling

    • Availability checking
    • Conflict detection
    • Automated reminders

Architecture​

Quick Start​

# Clone repository
git clone https://github.com/axonflow/examples
cd examples/healthcare-assistant

# Setup database
docker-compose up -d postgres

# Run backend
cd backend
go run main.go

# Run frontend
cd frontend
npm install
npm start

# Access at http://localhost:3000

Configuration​

Environment Variables:

# AxonFlow
AXONFLOW_ENDPOINT=https://your-agent-endpoint
AXONFLOW_LICENSE_KEY=AXON-V2-xxx-yyy
AXONFLOW_ORG_ID=healthcare-org

# Database
DATABASE_URL=postgresql://user:pass@localhost:5432/healthcare

# FHIR Integration (optional)
FHIR_BASE_URL=https://your-fhir-server
FHIR_CLIENT_ID=xxx
FHIR_CLIENT_SECRET=yyy

Policy Example​

package axonflow.policy.healthcare

import future.keywords

# HIPAA minimum necessary rule
allow {
    input.context.user_role in ["doctor", "nurse"]
    is_patient_assigned_to_user(
        input.context.user_id,
        extract_patient_id(input.query)
    )
}

# Emergency access (override with audit)
allow {
    input.context.emergency_access == true
    input.context.emergency_reason != ""
    log_emergency_access
}

# Block access to all patient data (HIPAA violation)
deny["HIPAA violation: minimum necessary rule"] {
    contains(lower(input.query), "all patients")
    input.context.user_role != "admin"
}

# Redact SSN from queries
redacted_query := regex.replace(
    input.query,
    `\b\d{3}-\d{2}-\d{4}\b`,
    "***-**-****"
)

Compliance Checklist​

  • ✅ Encryption at rest (AES-256)
  • ✅ Encryption in transit (TLS 1.3)
  • ✅ Access controls (RBAC)
  • ✅ Audit logging (all PHI access)
  • ✅ Data retention (6 years)
  • ✅ Breach notification procedures
  • ✅ Emergency access with audit trail
  • ✅ Business Associate Agreement (BAA) with AxonFlow

GitHub​

📁 View on GitHub

Documentation​

📖 Full Documentation

E-commerce Recommendations​

AI-powered product recommendation engine - Increase sales with personalized recommendations.

What It Demonstrates​

  • Product catalog integration
  • Personalized recommendations
  • Inventory management
  • Price calculation with policies
  • Cart management
  • Order processing workflow
  • A/B testing for recommendations

Features​

✅ Recommendations:

  • Collaborative filtering
  • Content-based filtering
  • Hybrid recommendations
  • Real-time personalization
  • Trending products

✅ Business Rules:

  • Dynamic pricing policies
  • Inventory constraints
  • Discount rules
  • Cross-sell/upsell logic
  • Geo-specific pricing

✅ Performance:

  • Sub-10ms recommendation latency
  • Real-time inventory checks
  • Cached product data
  • Batch processing for analytics

Use Cases​

  1. Product Page Recommendations

    • "Customers who bought this also bought..."
    • "Similar products you might like"
    • "Complete the look"

  2. Cart Recommendations

    • "Frequently bought together"
    • "Add these items to save more"
    • "Don't forget..."

  3. Personalized Homepage

    • Based on browsing history
    • Based on purchase history
    • Trending in your area

  4. Email Campaigns

    • Abandoned cart recovery
    • Product recommendations
    • Back-in-stock alerts

Architecture​

Quick Start​

cd examples/ecommerce-recommendations

# Setup
npm install
docker-compose up -d

# Run
npm run dev

# Access at http://localhost:3000

Policy Example​

package axonflow.policy.ecommerce

# Dynamic pricing based on user tier
apply_discount {
    input.context.user_tier == "premium"
    input.product.price_with_discount := input.product.price * 0.9
}

apply_discount {
    input.context.user_tier == "gold"
    input.product.price_with_discount := input.product.price * 0.85
}

# Geo-specific pricing
apply_geo_pricing {
    input.context.country == "US"
    input.product.price_usd := input.product.base_price
}

apply_geo_pricing {
    input.context.country == "EU"
    input.product.price_eur := input.product.base_price * 0.92
}

# Block out-of-stock recommendations
deny["Product out of stock"] {
    input.product.inventory_count <= 0
}

GitHub​

📁 View on GitHub

Documentation​

📖 Full Documentation

Customer Support​

Complete Community demo - AI governance for customer support with PII protection and RBAC.

Runnable Example: This is a fully-functional demo included in the Community repository. Clone and run with docker-compose up.

What It Demonstrates​

  • PII Detection & Redaction (SSNs, credit cards, phone numbers)
  • Role-Based Access Control (agents, managers, admins)
  • Policy Enforcement (SQL injection prevention)
  • Audit Logging (complete data access trail)
  • LLM Integration (natural language to SQL)

Features​

✅ Governance:

  • Automatic SSN/credit card redaction
  • Role-based PII visibility
  • SQL injection blocking
  • Dangerous query prevention

✅ RBAC:

  • Agent: Limited PII, regional access
  • Manager: Full PII, escalation handling
  • Admin: Global access, system admin

✅ Audit:

  • Complete query logging
  • User action tracking
  • Policy violation recording
  • Compliance reporting

Demo Scenarios​

  1. Agent Query (PII Redaction)

    • Login as support agent
    • Query customer data
    • See SSNs automatically redacted

  2. Manager Query (Full Access)

    • Login as manager
    • Query same data
    • See full PII (role-based)

  3. SQL Injection Prevention

    • Try malicious query: DROP TABLE users;
    • Query blocked by policy

Architecture​

┌─────────────────┐     ┌─────────────────┐     ┌─────────────────┐
│  React Frontend │────▶│   Go Backend    │────▶│   PostgreSQL    │
│   (Port 3000)   │     │   (Port 8080)   │     │   (Port 5432)   │
└─────────────────┘     └────────┬────────┘     └─────────────────┘
                                │
                                ▼
                       ┌─────────────────┐     ┌─────────────────┐
                       │ AxonFlow Policy │────▶│    LLM APIs     │
                       │    Engine       │     │ (OpenAI/Claude) │
                       └─────────────────┘     └─────────────────┘

Quick Start​

# Clone Community repository
git clone https://github.com/getaxonflow/axonflow.git
cd axonflow/platform/examples/support-demo

# Set API key
export OPENAI_API_KEY=sk-your-key
# OR
export ANTHROPIC_API_KEY=sk-ant-your-key

# Start
docker-compose up -d

# Access at http://localhost:3000

Demo Users​

EmailRolePassword
[email protected]Agentdemo123
[email protected]Managerdemo123
[email protected]Admindemo123

Policy Example​

package axonflow.policy.support

# Redact SSN for non-managers
redacted_query := regex.replace(
    input.query,
    `\b\d{3}-\d{2}-\d{4}\b`,
    "***-**-****"
) {
    input.context.user_role != "manager"
    input.context.user_role != "admin"
}

# Block SQL injection
deny["SQL injection blocked"] {
    contains(upper(input.query), "DROP TABLE")
}

deny["SQL injection blocked"] {
    contains(upper(input.query), "DELETE FROM")
}

GitHub​

📁 View on GitHub

Documentation​

📖 Full Documentation

Trip Planner​

AI-powered travel planning assistant - Multi-agent coordination with MCP connectors.

What It Demonstrates​

  • Multi-Agent Parallel (MAP) execution
  • MCP connector integration (Amadeus)
  • LLM integration (Claude)
  • Real-time flight/hotel search
  • Itinerary generation
  • Rate limiting patterns
  • Service identity & permissions

Features​

✅ Travel Planning:

  • Flight search (Amadeus API)
  • Hotel recommendations
  • Activity suggestions
  • Weather forecasts
  • Restaurant recommendations
  • Complete itinerary generation

✅ Performance:

  • 5x faster with parallel execution
  • Sub-10ms policy enforcement
  • Real-time availability
  • Cached results

✅ Advanced Patterns:

  • Service-based authentication
  • MCP connector permissions
  • Rate limiting (100/hour, 500/day)
  • Graceful fallback (LLM)
  • Multi-agent orchestration

Use Cases​

  1. Complete Trip Planning

    • Search flights
    • Find hotels
    • Suggest activities
    • Generate itinerary
    • All in single request

  2. Budget Optimization

    • Compare flight prices
    • Find best hotel deals
    • Optimize for budget
    • Suggest alternatives

  3. Multi-City Trips

    • Complex routing
    • Multiple destinations
    • Optimized connections
    • Time zone handling

Architecture​

Quick Start​

cd examples/trip-planner

# Setup backend
cd backend
go mod download
go run main.go

# Setup frontend
cd frontend
npm install
npm run dev

# Access at http://localhost:3000

Service License Configuration​

// Service identity with MCP permissions
const client = new AxonFlowClient({
  endpoint: process.env.AXONFLOW_ENDPOINT,
  licenseKey: process.env.SERVICE_LICENSE_KEY,  // Service-specific key
  organizationId: 'travel-agency',
  serviceIdentity: {
    name: 'trip-planner',
    type: 'backend-service',
    permissions: [
      'mcp:amadeus:search_flights',
      'mcp:amadeus:search_hotels',
      'mcp:amadeus:lookup_airport'
    ]
  }
});

Policy Example​

package axonflow.policy.travel

# Allow service to access Amadeus API
allow {
    input.service.name == "trip-planner"
    input.service.permissions[_] == sprintf("mcp:amadeus:%s", [input.mcp.operation])
}

# Enforce rate limiting
deny["Rate limit exceeded"] {
    request_count := count_requests_last_hour(input.service.name)
    request_count > 100
}

# Budget validation
deny["Budget too low"] {
    input.budget < 100
    input.budget_type == "total"
}

Performance​

Sequential Execution (traditional):

Flight search: 5s
Hotel search: 5s
Activities: 8s
Weather: 3s
Restaurants: 8s
Total: 29 seconds

Parallel Execution (MAP):

All 5 queries in parallel: 8s (max of all)
Speedup: 3.6x

GitHub​

📁 View on GitHub

Documentation​

📖 Full Documentation

Choosing the Right Example​

By Industry​

IndustryRecommended ExampleKey Features
Banking/FinanceBanking AI AssistantRBI, PCI-DSS, fraud detection, PII
HealthcareHealthcare AssistantHIPAA, PII protection, RBAC
Retail/E-commerceE-commerce RecommendationsPersonalization, inventory, pricing
SaaS/TechCustomer SupportTicket automation, knowledge base
TravelTrip PlannerMAP, MCP connectors, LLM

By Complexity​

Beginner: Start with Hello World, then try E-commerce or Customer Support

Intermediate: E-commerce Recommendations or Customer Support

Advanced: Banking AI Assistant, Healthcare Assistant, or Trip Planner (compliance, multi-agent, MCP)

By Features​

Need MCP Connectors? → Trip Planner or Healthcare Assistant

Need Multi-Agent (MAP)? → Trip Planner

Need HIPAA Compliance? → Healthcare Assistant

Need RBI/PCI-DSS Compliance? → Banking AI Assistant

Need Fraud Detection? → Banking AI Assistant

Need LLM Integration? → Trip Planner or Customer Support

Running Examples Locally​

Prerequisites​

All examples require:

  1. AxonFlow deployed (see Getting Started)
  2. License key (from CloudFormation outputs)
  3. Node.js 18+ or Go 1.21+

General Steps​

# 1. Clone repository
git clone https://github.com/axonflow/examples
cd examples/[example-name]

# 2. Configure environment
cp .env.example .env
# Edit .env with your credentials

# 3. Install dependencies
npm install  # or go mod download

# 4. Run
npm start    # or go run main.go

Contributing​

We welcome contributions! See CONTRIBUTING.md

Adding a New Example​

  1. Follow existing structure
  2. Include README with setup instructions
  3. Add .env.example file
  4. Include sample policies
  5. Add tests
  6. Update this overview page

Support​

Questions about examples?

All examples tested with AxonFlow v1.0.12 - Last updated: December 5, 2025