Enterprise Integration Guide

Complete guide for enterprise customers integrating with Mindra Platform.

Overview

Mindra Platform Enterprise provides:

• Dedicated Capacity: Reserved orchestration capacity on Mindra's infrastructure
• Bring Your Own Agents: Deploy agents on Mindra infra OR connect your external agent APIs
• Custom SLAs: Guaranteed uptime and performance (99.95% - 99.99%)
• Priority Support: Dedicated support team and Slack channel
• Volume Discounts: Reduced pricing for high-volume usage
• Private Agent Registry: Keep your internal agents private
• Advanced Security: SSO, SAML, audit logs, SOC 2 compliance
• Custom Integrations: Tailored workflows and API integrations

Getting Started

1. Contact Enterprise Sales

Email: info@mindra.co

Provide:

• Company name and size
• Expected monthly volume (executions)
• Use cases and verticals
• Custom requirements
• Timeline

2. Requirements Discovery

Work with our solutions architect to:

• Define technical requirements
• Plan integration architecture
• Identify custom needs
• Establish SLAs
• Set deployment timeline

3. Contract & Onboarding

• Review and sign Enterprise Agreement
• Provision dedicated infrastructure
• Configure deployment settings
• Set up SSO/SAML
• Create initial admin accounts
• Schedule training sessions

4. Integration & Testing

• Integrate with your systems
• Deploy custom agents
• Configure workflows
• Perform load testing
• Security audit
• User acceptance testing

5. Go Live

• Production cutover
• Monitor initial usage
• Address any issues
• Collect feedback
• Optimize performance

Enterprise Features

Dedicated Infrastructure

Included:

• Isolated Kubernetes cluster
• Dedicated databases (PostgreSQL, Redis)
• Private agent registry
• Dedicated monitoring stack (Prometheus, Grafana)
• Isolated networking (VPC)
• Custom capacity planning

Specifications (Typical):

• 10+ dedicated nodes
• 100+ GB RAM
• High-performance SSD storage
• Multi-AZ deployment
• Auto-scaling enabled

Custom SLAs

Standard Enterprise SLA:

• Uptime: 99.95% monthly
• API Response Time: P95 < 200ms
• Agent Execution Start: < 2 seconds
• Support Response:
  • Critical: < 1 hour
  • High: < 4 hours
  • Medium: < 24 hours
• Scheduled Maintenance: < 4 hours/month with 7-day notice

Premium SLA (Optional):

• Uptime: 99.99% monthly
• API Response Time: P95 < 100ms
• 24/7 Phone Support
• Dedicated Solutions Architect
• Zero-downtime deployments

Private Agent Registry

Deploy internal agents not visible to other customers.

# Register private agent
curl -X POST https://api.your-company.mindra.co/v1/agents \
  -H "Authorization: Bearer mk_your_enterprise_key" \
  -d '{
    "id": "internal-hr-agent",
    "name": "Internal HR Agent",
    "url": "https://internal.acme.com/hr-agent",
    "visibility": "private",  // Only visible to your org
    "vertical": "hr"
  }'

Advanced Security

Single Sign-On (SSO):

• SAML 2.0
• OAuth 2.0 / OpenID Connect
• Active Directory integration
• Okta, Azure AD, Google Workspace

Access Control:

• Role-Based Access Control (RBAC)
• Custom permission policies
• IP whitelisting
• API key expiration policies

Compliance:

• SOC 2 Type II certified
• GDPR compliant
• HIPAA available (Healthcare plan)
• ISO 27001 certified
• Regular penetration testing

Audit Logs:

# Query audit logs
curl https://api.your-company.mindra.co/v1/audit-logs \
  -H "Authorization: Bearer mk_admin_key" \
  -d '{
    "startDate": "2025-11-01",
    "endDate": "2025-11-10",
    "eventTypes": ["agent.registered", "execution.started"]
  }'

Integration Patterns

Pattern 1: API Gateway

Use Mindra as your central AI orchestration layer.

┌──────────────┐
│  Your App    │
│  (Frontend)  │
└──────┬───────┘
       │
       ▼
┌──────────────────────────┐
│   Your API Gateway       │
│   - Auth                 │
│   - Rate Limiting        │
│   - Logging              │
└──────┬───────────────────┘
       │
       ▼
┌──────────────────────────┐
│   Mindra Platform        │
│   - Agent Orchestration  │
│   - Workflow Engine      │
└──────┬───────────────────┘
       │
       ├──► Your Custom Agents
       ├──► Third-party Agents
       └──► Mindra Agents

Pattern 2: Embedded Workflows

Embed Mindra workflows directly in your application.

import { Mindra } from '@mindra/sdk';

const mindra = new Mindra({
  apiKey: process.env.MINDRA_API_KEY,
  baseUrl: 'https://api.your-company.mindra.co'
});

// In your application code
async function processOrder(order) {
  // Use Mindra for complex AI workflows
  const result = await mindra.orchestrate({
    query: `Process order ${order.id}: validate inventory, check fraud, arrange shipping`,
    context: { order },
    mode: 'DAG'
  });

  return result;
}

Pattern 3: Event-Driven

Trigger Mindra orchestrations from events.

// Your event handler
async function handleOrderCreated(event) {
  // Trigger async Mindra workflow
  await mindra.orchestrate({
    query: 'Send order confirmation email and update CRM',
    context: { orderId: event.orderId },
    webhook: {
      url: 'https://your-app.com/webhooks/mindra',
      events: ['execution.completed']
    }
  });
}

// Your webhook handler
app.post('/webhooks/mindra', async (req, res) => {
  const { event, data } = req.body;

  if (event === 'execution.completed') {
    // Update your database
    await db.orders.update({
      where: { id: data.context.orderId },
      data: { aiProcessed: true, result: data.result }
    });
  }

  res.json({ ok: true });
});

Security Setup

SSO Configuration (SAML)

<!-- SAML Configuration -->
<EntityDescriptor xmlns="urn:oasis:names:tc:SAML:2.0:metadata"
                  entityID="https://api.acme-ai.com/saml/metadata">
  <SPSSODescriptor protocolSupportEnumeration="urn:oasis:names:tc:SAML:2.0:protocol">
    <AssertionConsumerService
      Binding="urn:oasis:names:tc:SAML:2.0:bindings:HTTP-POST"
      Location="https://api.acme-ai.com/saml/acs"
      index="0"/>
  </SPSSODescriptor>
</EntityDescriptor>

IP Whitelisting

curl -X POST https://api.your-company.mindra.co/v1/security/ip-whitelist \
  -H "Authorization: Bearer mk_admin_key" \
  -d '{
    "cidrs": [
      "203.0.113.0/24",
      "198.51.100.42/32"
    ],
    "description": "Corporate office IPs"
  }'

Role-Based Access Control

# Create custom role
curl -X POST https://api.your-company.mindra.co/v1/roles \
  -H "Authorization: Bearer mk_admin_key" \
  -d '{
    "name": "Agent Developer",
    "permissions": [
      "agents:read",
      "agents:write",
      "executions:read"
    ]
  }'

# Assign role to user
curl -X POST https://api.your-company.mindra.co/v1/users/user_123/roles \
  -H "Authorization: Bearer mk_admin_key" \
  -d '{
    "roleId": "role_agent_developer"
  }'

Agent Deployment Options

You have two options for deploying your custom agents:

Option 1: Deploy on Mindra Infrastructure (Recommended)

Deploy your agents directly on Mindra's infrastructure for seamless integration.

Benefits:

• Fast deployment (< 1 hour)
• Managed infrastructure and scaling
• Automatic health checks and monitoring
• Low latency to orchestrator
• No infrastructure management needed

How it works:

1. Package your agent as Docker container
2. Push to Mindra's private registry
3. Configure via dashboard or API
4. Mindra handles deployment, scaling, and monitoring

Example:

# Build and push your agent
docker build -t my-custom-agent .
docker tag my-custom-agent registry.mindra.co/your-org/my-custom-agent
docker push registry.mindra.co/your-org/my-custom-agent

# Register via API
curl -X POST https://api.mindra.co/v1/agents \
  -H "Authorization: Bearer mk_your_key" \
  -d '{
    "name": "my-custom-agent",
    "image": "registry.mindra.co/your-org/my-custom-agent:latest",
    "visibility": "private"
  }'

Option 2: Connect External Agent APIs

Keep your agents on your own infrastructure and connect them via HTTP API.

Benefits:

• Full control over agent infrastructure
• Data stays in your environment
• Use existing infrastructure
• Flexible deployment (AWS, GCP, Azure, on-prem)

Requirements:

• Agent must expose HTTP API endpoints
• Must be accessible from Mindra's infrastructure
• Follow Mindra Agent Protocol specification

Example:

# Register external agent
curl -X POST https://api.mindra.co/v1/agents \
  -H "Authorization: Bearer mk_your_key" \
  -d '{
    "name": "my-external-agent",
    "url": "https://your-agent.your-company.com",
    "visibility": "private",
    "authMethod": "bearer",
    "authToken": "your_agent_secret"
  }'

Architecture:

┌─────────────────────────┐
│   Mindra Platform       │
│   (SaaS)                │
│   - Orchestrator        │
│   - Agent Registry      │
└───────────┬─────────────┘
            │
            ├──► Agents on Mindra Infra
            │    (Option 1)
            │
            └──► Your Infrastructure
                 ├── AWS/GCP/Azure
                 │   └── Your Agent APIs
                 └── On-Premises
                     └── Your Agent APIs

Best Practices:

• Use API authentication (Bearer token, API key)
• Implement proper error handling
• Return consistent response formats
• Monitor agent health and availability
• Use HTTPS for all communication

Pricing

Enterprise Tiers

Feature	Standard	Premium	Ultimate
Price	Custom	Custom	Custom
Executions	Up to 1M/month	Up to 10M/month	Unlimited
SLA	99.95%	99.99%	99.995%
Support	Business hours	24/7	24/7 + Phone
Dedicated Capacity	✓	✓	✓
Agent Hosting	✓	✓	✓
Private Agents	✓	✓	✓
SSO/SAML	✓	✓	✓
Custom SLA	-	✓	✓
Dedicated SA	-	✓	✓
Volume Discounts	-	✓	✓

Volume Discounts

Monthly Volume	Discount
100K executions	0%
500K executions	10%
1M executions	20%
5M executions	30%
10M+ executions	35%+ (negotiable)

Support

Enterprise Support Channels

Email: info@mindra.co

• Response Time: < 1 hour for critical issues

Dedicated Slack Channel:

• Direct access to engineering team
• Shared channel with your team
• Real-time support

Phone Support (Premium/Ultimate):

• +1 (XXX) XXX-XXXX
• 24/7 availability

Solutions Architect (Premium/Ultimate):

• Quarterly business reviews
• Architecture guidance
• Performance optimization
• Capacity planning

Escalation Path

1. Level 1: Support engineer (< 1 hour)
2. Level 2: Senior engineer (< 2 hours)
3. Level 3: Engineering manager (< 4 hours)
4. Level 4: CTO (critical incidents)

Migration Support

Migrating from another platform? We provide:

• Migration Planning: Detailed migration strategy
• Data Migration: Tools and assistance for data transfer
• Code Migration: Help adapting your code
• Testing Support: Comprehensive testing assistance
• Cutover Support: Hands-on support during go-live
• Post-Migration: 30 days of enhanced support

Training & Onboarding

Training Programs

Developer Training (2 days):

• API fundamentals
• Agent development
• Best practices
• Hands-on labs

Administrator Training (1 day):

• Platform management
• User administration
• Monitoring and alerting
• Security configuration

Architecture Training (1 day):

• Platform architecture
• Integration patterns
• Performance optimization
• Troubleshooting

Documentation

• Private documentation portal
• Custom integration guides
• Video tutorials
• Code examples for your use cases

Getting Started

Ready to integrate Mindra Platform?

1. Contact Enterprise Sales: info@mindra.co
2. Schedule Demo: See platform in action
3. Proof of Concept: 30-day trial with support
4. Sign Contract: Enterprise Agreement
5. Onboarding: Dedicated onboarding team

FAQs

Q: Can we deploy agents using any technology? A: Yes! Python, Go, TypeScript, Rust, Java, or any language that can run an HTTP server.

Q: Can we use our own ML models instead of LLMs? A: Absolutely! Agents can run any logic - ML models, databases, APIs, blockchain, etc.

Q: Do we need to share our agent code with Mindra? A: No. Your agents run on your infrastructure. We only need the HTTP endpoint.

Q: What's the minimum contract term? A: 12 months for Standard/Premium, negotiable for Ultimate.

Q: Can we start small and scale up? A: Yes! Start with Standard tier and upgrade as you grow.

Q: Is there a free trial? A: Yes, 30-day proof of concept with full Enterprise features.

---

Contact: info@mindra.co | Schedule Demo: calendly.com/mindra-enterprise