# ozIQ Architecture Overview

> Investor-level technical overview. This document is intentionally high level and is not a production security specification.

## What ozIQ Connects To

ozIQ is designed to connect to enterprise BI platforms through customer-authorized, read-only interfaces.

- **Initial platform:** Strategy
- **Expansion platforms:** Power BI and Tableau
- **Long-term direction:** A cross-platform operational intelligence layer for enterprise analytics

## High-Level Data Flow

```text
Enterprise BI Platforms
Strategy → Power BI → Tableau
                │
                ▼
Customer-Authorized Read-Only Collection
                │
                ▼
Secure Cloud Ingestion and Processing
                │
                ▼
Operational Telemetry Storage
                │
                ▼
ozIQ Intelligence Layer
Health • Workloads • Alerts • Recommendations • AI Advisory
                │
                ▼
BI Leaders • Platform Administrators • Operations Teams
```

## What ozIQ Collects

ozIQ focuses on operational telemetry needed to understand the health and behavior of enterprise analytics environments, including:

- Job and execution activity
- Runtime and queue behavior
- Failures and operational events
- Cache and workload telemetry
- Schedule and subscription activity
- Configuration and change signals
- Platform usage and adoption indicators

## What ozIQ Does Not Collect

- Business report data
- Dashboard results
- Customer source-system records
- Underlying transactional or analytical datasets

This separation allows ozIQ to provide operational intelligence without becoming a repository for customer business data.

## Core Trust Principles

- Customer-provided, read-only platform access
- Telemetry-only collection
- PII masked by default
- Tenant-aware access controls
- Encryption in transit and at rest
- Secrets managed outside application code
- Customer-controlled connector configuration
- Architecture designed to support SOC 2 readiness

## Platform Architecture

The product uses a cloud-native architecture with separate layers for collection, processing, telemetry storage, configuration, and application access.

- **Collection layer:** Retrieves authorized operational telemetry from BI platforms
- **Processing layer:** Normalizes platform-specific events into a consistent operational model
- **Telemetry layer:** Stores time-series and event data for fast analysis
- **Configuration layer:** Stores customer, environment, access, and connector settings
- **Application layer:** Presents health, workload, alert, recommendation, and advisory experiences

## Why the Architecture Matters

The architecture supports three important parts of the investment thesis:

1. **Focused entry:** ozIQ can build deep Strategy-specific value first.
2. **Reusable platform:** The operational data model and intelligence layer can be extended to Power BI and Tableau.
3. **Enterprise fit:** Telemetry-only collection and read-only access reduce adoption and data-governance friction.

## Scalability Approach

The system is designed so that collection and processing workloads can scale independently by customer, environment, and platform. Telemetry storage is optimized for high-volume operational events and historical trend analysis.

The initial product refresh target is approximately every 15 minutes, with more frequent collection possible for selected running-job use cases.

## Technical Diligence

Detailed infrastructure diagrams, deployment topology, data models, controls, and security documentation can be provided during technical diligence under the appropriate confidentiality process.