Enterprise Provider Features

Enterprise Feature

Runtime provider management via Customer Portal is available in AxonFlow Enterprise Edition.

Contact Sales | Request Demo

Overview

Enterprise customers get additional LLM provider capabilities through the Customer Portal:

Feature	Community	Enterprise
All LLM Providers (OpenAI, Azure OpenAI, Anthropic, Bedrock, Gemini, Ollama)	✅	✅
Multi-Provider Routing	✅	✅
Automatic Failover	✅	✅
Circuit Breaker	✅	✅
YAML Configuration	✅	✅
Customer Portal UI	❌	✅
Runtime Configuration	❌	✅
Secure Credential Storage	❌	✅
Per-Provider Metrics	❌	✅
Cost Tracking Dashboard	❌	✅
API Key Rotation	❌	✅

Customer Portal UI

Enterprise customers can manage LLM providers through a web interface:

Provider Configuration

• Add/remove providers without code changes
• Update API keys securely
• Adjust routing weights in real-time
• Changes take effect within 30 seconds

Real-Time Monitoring

• Per-provider request counts
• Latency percentiles (P50, P95, P99)
• Error rates and types
• Cost per provider/model

Credential Management

• Secure API key storage (encrypted at rest)
• Key rotation without downtime
• Audit log for credential changes
• Role-based access control

Runtime Configuration

Change provider settings without redeployment:

Via Customer Portal API

# Update provider weight
curl -X PATCH https://your-axonflow.example.com/v1/providers/openai \
  -H "Authorization: Bearer $AXONFLOW_API_KEY" \
  -d '{"weight": 0.3}'

# Disable a provider temporarily
curl -X PATCH https://your-axonflow.example.com/v1/providers/anthropic \
  -H "Authorization: Bearer $AXONFLOW_API_KEY" \
  -d '{"enabled": false}'

# Update routing strategy
curl -X PUT https://your-axonflow.example.com/v1/routing \
  -H "Authorization: Bearer $AXONFLOW_API_KEY" \
  -d '{"strategy": "cost-optimized"}'

Changes Propagate Automatically

1. Configuration change saved to Customer Portal
2. Orchestrator polls for changes (30-second interval)
3. New configuration applied atomically
4. No restarts or downtime required

Cost Tracking

Enterprise includes detailed cost analytics:

Per-Provider Costs

Provider	Requests	Tokens	Cost
OpenAI	45,000	12.3M	$234.56
Anthropic	32,000	8.1M	$123.45
Bedrock	23,000	5.6M	$78.90

Cost Optimization Recommendations

The Customer Portal analyzes usage patterns and provides recommendations:

• "Route 20% more traffic to Bedrock to save $50/month"
• "Enable Ollama for development queries to reduce costs"
• "Consider Llama 3.1 for simple completions"

Credential Rotation

Enterprise customers can rotate LLM API keys without downtime using the Customer Portal or API.

Rotation Process

1. Add new key -- upload the new API key via the Portal or API. The new key is stored encrypted alongside the current key.
2. Validate -- AxonFlow sends a test request using the new key to confirm it works.
3. Activate -- once validated, the new key becomes the active key. This happens atomically on the next provider poll (within 30 seconds).
4. Revoke old key -- after confirming traffic is flowing on the new key, revoke the old key in the provider's console.

Via API

# Step 1: Upload new key (does not activate yet)
curl -X POST https://your-axonflow.example.com/v1/providers/openai/credentials/rotate \
  -H "Authorization: Bearer $AXONFLOW_API_KEY" \
  -d '{"new_api_key": "sk-new-key-here"}'

# Step 2: Activate the new key
curl -X POST https://your-axonflow.example.com/v1/providers/openai/credentials/activate \
  -H "Authorization: Bearer $AXONFLOW_API_KEY"

Audit Logging for Provider Operations

All credential and provider configuration changes are recorded in the audit trail:

Event	Description
provider.credential.rotated	New API key uploaded
provider.credential.activated	New key set as active
provider.config.updated	Weight, priority, or model changed
provider.enabled	Provider enabled
provider.disabled	Provider temporarily disabled

Audit entries include the user who made the change, timestamp, and the previous/new values (keys are masked).

SLA Management

Set per-provider SLOs:

# Enterprise configuration
providers:
  openai:
    slo:
      latency_p99: 5s
      error_rate: 0.1%
      availability: 99.9%
    alerts:
      - type: latency_threshold
        threshold: 5s
        channel: pagerduty

Alerting Integration

• PagerDuty
• Slack
• OpsGenie
• Custom webhooks

Getting Started

Step 1: Upgrade to Enterprise

Contact our sales team:

• Email: [email protected]
• Demo: getaxonflow.com/demo

Step 2: Access Customer Portal

After upgrading:

1. Log in to app.getaxonflow.com
2. Navigate to Settings > LLM Providers
3. Configure providers via the UI

Step 3: Migrate from YAML (Optional)

Existing YAML configurations can be imported:

axonctl providers import --file axonflow.yaml

Pricing

Enterprise LLM features are included in all paid tiers:

SaaS (AxonFlow-Hosted)

Tier	Monthly	Included Requests
Starter	$5,000	500K/month
Professional	$15,000	3M/month
Enterprise	$50,000	10M/month

In-VPC (Self-Hosted)

Tier	Monthly	Max Nodes
Professional	$20,000	10
Enterprise	$60,000	50
Enterprise Plus	Custom	Unlimited

Next Steps

• LLM Providers Overview - All supported providers
• AWS Bedrock Setup - HIPAA-compliant deployment
• Custom Provider SDK - Build custom providers