Haultro Security. Enterprise-grade security architecture for waste management software. Haultro by Viceroy NM provides multi-tenant data isolation at the database level with organization-scoped queries on every request. JWT-based authentication with embedded role claims enforces 5 distinct permission levels (Admin, Manager, Driver, Client, Enterprise Admin) across 150 plus API endpoints. WebSocket real-time connections use Socket.IO room-based architecture scoped to organization ID for complete event isolation. AI safety guardrails include input sanitization, iteration limits, and role-based tool restrictions for the Haultro AI Chat. IoT sensor endpoint authenticates device identity before accepting readings. All data in transit encrypted with TLS and HTTPS. PostgreSQL database with 20 plus relational tables and organization-level data isolation. Redis for session management and high-throughput IoT ingestion buffering. Containerized deployment with 7 interconnected services. Stripe PCI-compliant payment processing. Haultro never stores credit card data. AI provider communications with Anthropic Claude, OpenAI, and Google use encrypted channels under commercial API terms with no customer data used for model training. Platform-wide audit logging captures every action across every organization for compliance-grade audit trails.

Security & Trust

Your data is your business. We protect it.

Multi-tenant data isolation. Role-based access control. Encrypted communications. Platform-wide audit logging. Security built into the architecture from day one, not bolted on after launch.

5
Permission roles
TLS
Encryption in transit
7
Isolated services
24/7
Audit logging
The Architecture

Five layers between a request and your data.

Every API call passes through five sequential security checks before it ever touches your data. Watch a request flow from edge to database below.

REQUEST User or Device TLS / HTTPS Encrypt in transit LAYER 1 JWT AUTH Verify identity LAYER 2 ROLE CHECK 5 permission tiers LAYER 3 ORG SCOPE Tenant isolation LAYER 4 DATABASE Your data LAYER 5 EVERY ACTION LOGGED FOR AUDIT
Security Pillars

Three pillars hold up the platform.

Data isolation. Access control. Audit and compliance. Each pillar is enforced at multiple layers, from the database query up to the API edge.

Data Isolation

Complete tenant isolation at the database level. Every query is scoped to the authenticated user's organization. Cross-organization data access is architecturally impossible for standard users. Only authenticated Enterprise Admins can perform cross-org queries, and only through the Portfolio Intelligence module.

Access Control

JWT-based authentication with role claims enforced on every API request. Five distinct permission levels (Admin, Manager, Driver, Client, Enterprise Admin) each with precisely scoped access to endpoints, data, and features. 150 plus endpoints all enforce org membership and role validation.

Audit and Compliance

Platform-wide audit logging captures every action across every organization. User actions, API calls, configuration changes, and data access are all timestamped and immutable. Enterprise operators get compliance-grade audit trails exportable for regulatory reviews and internal governance.

Authentication

Identity and access management.

Every request to the Haultro platform is authenticated, authorized, and scoped. There are no anonymous endpoints for operational data. The security model validates three layers on every API call: identity (who you are), organization (which tenant), and role (what you are allowed to do).

JWT token authentication
Signed JSON Web Tokens with embedded role claims, org membership, and expiration. Tokens validated on every request with no server-side session dependency.
5-role permission model
Admin, Manager, Driver, Client, Enterprise Admin. Drivers cannot access financials. Clients cannot see other clients' data. Permissions enforced server-side.
Organization-scoped queries
Every database query includes an org_id filter. Even if a user obtains another org's record ID, the query returns nothing. Isolation is at the data layer, not just the API layer.
Real-Time Security

WebSocket and API protection.

Real-time features use WebSocket connections with room-based architecture. Users are automatically placed in their organization's room and only receive events from that room. The AI Chat includes additional guardrails to prevent prompt injection and system abuse.

Room-based WebSocket isolation
Socket.IO rooms scoped to organization ID. A user in Org A never receives events from Org B even if both are online simultaneously.
AI safety guardrails
Haultro AI Chat includes input sanitization, iteration limits, and role-based tool restrictions. Drivers get read-only AI access. They can ask questions but cannot trigger system actions.
IoT endpoint security
The high-throughput sensor ingestion endpoint authenticates device identity before accepting readings. Invalid or unauthorized sensor data is rejected before it enters the data pipeline.
Infrastructure

Production-grade infrastructure.

The technology stack underneath every feature. Each service is isolated, encrypted, and monitored. Each integration is hardened.

PostgreSQL

20+ relational tables with org-level isolation. Foreign keys and indexes enforce data integrity end-to-end.

Redis

In-memory caching for sessions, real-time data layers, and high-throughput IoT ingestion buffering.

Socket.IO

WebSocket connections with room-based architecture for org-scoped real-time events. Auto reconnection and fallback.

HTTPS / TLS

All data in transit encrypted with TLS. API endpoints, WebSocket connections, and client portals all over HTTPS.

Containerized

Fully containerized deployment with 7 interconnected services. Isolated runtime environments for API, workers, and real-time.

Stripe Payments

Payment processing via Stripe with PCI-compliant infrastructure. Haultro never stores credit card data.

AI Provider Security

Encrypted channels to Anthropic, OpenAI, and Google under commercial API terms. No customer data used for model training.

Audit Logging

Every action across every org logged with timestamp, user ID, action type, and affected resources. Exportable for compliance.

Frequently Asked

Security questions, answered.

What customers ask before they trust us with their operations.

Is my organization's data isolated from other customers?
Yes, completely. Haultro uses multi-tenant architecture with organization-level data isolation enforced at the database layer. Every query includes an org_id filter. Your data cannot be accessed by users from other organizations. Even API requests with valid authentication are rejected if they reference data outside the authenticated user's organization. Cross-organization access is only available to Enterprise Admin users through the Portfolio Intelligence module, and only across organizations they have explicitly linked.
How does role-based access control work?
Haultro uses 5 user roles, each with precisely defined permissions enforced on all 150 plus API endpoints. Admins have full platform access including financials and system configuration. Managers can handle routes, fleet, drivers, and clients but cannot access financial analytics or system settings. Drivers only see their assigned routes and have read-only AI chat access. Clients can only view their own bins, request pickups, and review their contracts. Enterprise Admins can manage multiple organizations with cross-org analytics. Permissions are embedded in JWT tokens and validated on every request.
Is data encrypted in transit and at rest?
All data in transit is encrypted via TLS / HTTPS. This includes API calls, WebSocket connections, client portal access, and IoT sensor data ingestion. Database storage uses PostgreSQL's built-in security features with access restricted to the application layer only. Payment data is handled entirely by Stripe's PCI-compliant infrastructure. Haultro never stores, processes, or transmits credit card numbers. AI provider communications (Anthropic, OpenAI, Google) use encrypted API channels with provider-managed key security.
Does Haultro use my data to train AI models?
No. Your operational data is never used to train third-party AI models. Haultro accesses Anthropic Claude, OpenAI, and Google AI through their commercial API tiers, which by default exclude customer data from model training. When data is sent to these providers for features like fill predictions, intelligence reports, or Vision AI analysis, the data is processed for your specific request only. Your bin data, client information, routes, and financial data remain yours. They are not shared with other customers or used to improve general AI models.
What happens to my data if I cancel my subscription?
If you cancel, your data is retained for 30 days in case you want to reactivate. During this period, you can request a full data export in standard formats. After the 30-day retention period, all organization data (bins, routes, clients, sensor readings, reports, and user accounts) is permanently deleted from our systems. We provide confirmation of data deletion upon request. Enterprise and Portfolio tier customers can negotiate custom data retention terms as part of their service agreement.
Can I export my data?
Yes. Haultro supports data export at multiple levels. The platform includes one-click Excel exports for client reports, route analytics, and financial data. The API (150 plus RESTful endpoints) provides programmatic access to all your operational data (bins, routes, drivers, clients, sensor readings, alerts, and predictions) in standard JSON format. Enterprise and Portfolio tier customers have full API access for custom integrations and data warehousing. Your data is yours and you can extract it at any time.
Enterprise Security Reviews Welcome

Have specific security requirements?

Our team can walk you through the platform's security architecture in detail, answer compliance questions, and discuss custom security requirements for enterprise deployments.

Talk to Security Team Enterprise Solutions