Infrastructure Sizing & Scaling

AI Fabrix is designed to scale predictably inside the customer’s Azure tenant. Every deployment uses Azure-native resources, with defined size presets (S, M, L, XL) to align infrastructure with expected workloads and compliance needs.

This approach ensures that scaling is transparent, cost-predictable, and managed entirely within the customer’s Azure subscription.

Table of Contents

1. Sizing Profiles
2. Scaling Approach
3. Upgrade Path
4. Observability & Cost Control
5. Key Takeaway

Sizing Profiles

Each subscription is deployed with a size preset that determines the underlying Azure services for compute, database, caching, storage, and networking.

⚠️ Exact Azure SKUs may vary depending on regional availability and customer preferences. All infrastructure costs are billed directly to the customer’s Azure subscription for full transparency.

Scaling Approach

Fabrix uses Azure-native scaling options to adapt to enterprise workloads:

• Horizontal scaling — Container workloads and App Services autoscale based on demand.
• Vertical scaling — Upgrade from S → M → L → XL by adjusting Azure service plans.
• Multi-region distribution — Enabled with Azure Front Door for global, low-latency deployments.
• Environment lifecycle — Dev → Test → Prod separation ensures controlled promotion.
• Rollback support — Application containers can revert to a known-good state.

Upgrade Path

• Subscription Upgrades: Move from Community → Standard → Enterprise without redeploying infrastructure; features unlock via license.
• Size Upgrades: Move between S, M, L, XL with seamless migration; no architectural redesign required.

Observability & Cost Control

Fabrix adds governance and visibility on top of Azure monitoring:

• Telemetry: Metrics, structured logs, traces, and health endpoints.
• Cost Transparency: Direct Azure billing, no hidden SaaS fees.
• Quotas & Policies: Administrators can set usage and cost limits at the Miso control layer.

Infrastructure Scaling Diagram

flowchart LR
    S[S — Sandbox / Dev] --> M[M — SMB / Dept Scale] --> L[L — Enterprise Scale] --> XL[XL — Global / Regulated]

    S --- S1[Basic App Service / Small Container]
    S --- S2[PostgreSQL ~50 GB, Single-node]
    S --- S3[Redis Basic 1 GB]
    S --- S4[Storage 100 GB]
    S --- S5[VNet + Private Endpoints]

    M --- M1[Standard App Service]
    M --- M2[PostgreSQL HA ~200 GB]
    M --- M3[Redis 2–5 GB]
    M --- M4[Storage 500 GB]
    M --- M5[VNet + Private Endpoints]

    L --- L1[Premium App Service]
    L --- L2[PostgreSQL HA ~500 GB]
    L --- L3[Redis Premium 10 GB]
    L --- L4[Storage 1 TB]
    L --- L5[VNet + Private Endpoints]

    XL --- XL1[Premium/Isolated App Service]
    XL --- XL2[PostgreSQL >1 TB, HA]
    XL --- XL3[Redis Premium >20 GB]
    XL --- XL4[Storage 5 TB+]
    XL --- XL5[VNet + Global Load Balancing (Front Door)]

    classDef stage fill:#f3f6fa,stroke:#2c3e50,stroke-width:1px;
    class S,M,L,XL stage;

This diagram shows the progression from S → M → L → XL with specific infrastructure components and capabilities at each size tier.

Key Takeaway

Fabrix infrastructure sizing makes it easy to start small, grow to enterprise production, and scale globally — all while staying inside the same Azure-native architecture.