Integrations Overview

LangGuard integrates with leading AI observability and data platforms to provide unified governance and monitoring.

Supported Integrations

Observability Platforms

Platform	Capabilities	Status
Langfuse	Traces, Metrics	✓ Stable
Databricks MLflow	Traces, Data Catalog	✓ Stable
LangSmith	Traces, Runs	✓ Stable
Braintrust	Traces, Evals	✓ Stable
Helicone	Traces, Analytics	✓ Stable
Phoenix	Traces, Spans	✓ Stable
Weave	Traces, Artifacts	✓ Stable

Identity & Access

Platform	Capabilities	Status
Okta	SSO, User Sync	✓ Stable

Architecture

LangGuard uses a unified integration architecture:

┌────────────────────────────────────────────────────────────┐
│                      LangGuard Dashboard                   │
├────────────────────────────────────────────────────────────┤
│                    Sync Orchestrator                       │
├────────────────────────────────────────────────────────────┤
│                   Connector Registry                       │
├──────────┬──────────┬──────────┬──────────┬────────────────┤
│ Langfuse │Databricks│LangSmith │  Okta    │    Others      │
└──────────┴──────────┴──────────┴──────────┴────────────────┘
                          │
            ┌─────────────┼─────────────┐
            ▼             ▼             ▼
      ┌───────────┐ ┌───────────┐ ┌───────────┐
      │ Langfuse  │ │Databricks │ │ LangSmith │
      │   Cloud   │ │   MLflow  │ │   Cloud   │
      └───────────┘ └───────────┘ └───────────┘

Key Features

• Unified Credential Storage - AES-256-GCM encrypted credentials
• Automatic Sync - Configurable polling intervals
• Push-Based Ingestion - OpenTelemetry webhook support
• Source Tracking - Visual indicators for data origin
• Extensible - Easy to add new integrations

Adding an Integration

Via UI

1. Navigate to Settings > Integrations
2. Click Add Integration
3. Select the platform
4. Enter credentials
5. Test connection
6. Configure sync settings
7. Save

Via API

curl -X POST "https://api.langguard.ai/v1/integrations" \
  -H "Authorization: Bearer $API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "type": "langfuse",
    "name": "Production Langfuse",
    "credentials": {
      "publicKey": "pk-lf-...",
      "secretKey": "sk-lf-..."
    }
  }'

Credential Security

Encryption

All credentials are encrypted before storage:

• Algorithm: AES-256-GCM
• Key: 256-bit key from ENCRYPTION_KEY env var
• Storage: Only encrypted data in database
• Transit: HTTPS for all API calls

Best Practices

1. Use service accounts - Don't use personal API keys
2. Rotate regularly - Update keys quarterly
3. Least privilege - Only grant needed permissions
4. Monitor access - Review API key usage logs

Sync Configuration

Automatic Sync

Configure automatic synchronization:

Setting	Description	Default
Enabled	Auto-sync on/off	Yes
Interval	Time between syncs	5 minutes
Lookback	Days of history	7 days
Batch Size	Items per sync	100

Manual Sync

Trigger sync on-demand:

• Click Sync Now button
• Or call POST /api/integrations/{id}/sync

Push-Based Ingestion

For real-time ingestion, use OpenTelemetry webhooks:

POST /api/webhooks/opentelemetry
Headers:
  x-tenant-slug: your-tenant
  x-otel-signature: sha256=<hmac-signature>

See individual integration guides for setup.

Monitoring Integration Health

Connection Status

Status	Meaning
🟢 Connected	Credentials valid, sync working
🟡 Warning	Minor issues (rate limits, etc.)
🔴 Failed	Connection or auth failed
⚪ Disabled	Integration manually disabled

Sync History

View recent sync operations:

• Success/failure status
• Items fetched and ingested
• Duration
• Error messages (if any)

Troubleshooting

Common issues and solutions:

1. Auth Failed - Check credentials are correct
2. Rate Limited - Increase sync interval
3. Timeout - Reduce batch size
4. No Data - Check time range and filters

See Integration Issues for detailed help.

Quick Links

Langfuse

LLM observability platform with traces and metrics.

Databricks

MLflow integration for ML traces and Unity Catalog.

LangSmith

LangChain's observability and debugging platform.

Okta

Enterprise SSO and user directory sync.

Braintrust

AI evaluation and tracing platform.

Helicone

LLM analytics and cost tracking.

Phoenix

Arize AI's open-source observability.

Weave

Weights & Biases ML observability.

---

Need a New Integration?

Don't see your platform? Request an integration or check our developer guide to build your own connector.