EMCI — Enterprise Multi-Cloud Complexity Index

Organisation profile

Provide context for peer benchmarking and industry-specific findings.

Organisation Name ?The name of your company or business unit. Used in the report header and confidentiality notice.

Industry ?Your primary industry vertical. This determines which peer benchmark cohort your score is compared against.

Annual Cloud Spend ?Your total annual spend across all cloud providers combined. Used to estimate networking waste and ROI of remediation.

Revenue per Hour of Downtime ?Estimated revenue lost per hour when critical infrastructure is unavailable. Used to calculate financial exposure from unplanned outages.

Assessment Prepared For ?Name and title of the person who will receive this report. Appears in the report header for personalisation.

Infrastructure domains

Select all cloud providers and on-prem infrastructure in your environment.

Cloud Providers ?All public cloud providers your organisation actively uses for production or staging workloads. Each provider adds management overhead and inter-cloud connectivity requirements.

AWS

Azure

GCP

OCI

Other CSP

Select at least one cloud provider.

info

Hybrid Infrastructure: Private data centers, colocations, and edge sites are treated as additional infrastructure domains. Each adds management surface area and cross-domain connectivity requirements.

Private Data Centers ?The number of private data centers your organisation owns or operates. These add distinct management planes, physical security requirements, and connectivity overhead.

Colocation Facilities ?Third-party data center facilities where you rent space and power. Each adds a unique vendor relationship and interconnection architecture.

Edge / Branch Sites ?Remote locations with compute or networking equipment — branch offices, retail stores, factory floors, cell towers, etc. Complexity scales logarithmically with edge count.

Private Cloud Platform ?If you run a private cloud platform, it automatically adds its associated management tools to your toolchain count (e.g., VMware adds vCenter, NSX Manager, vRealise, etc.).

None

VMware vSphere/NSX

Nutanix AHV

OpenStack

Azure Stack HCI

Proxmox

Network topology

Cloud regions, accounts, network segments, and interconnections.

Cloud Regions ?Total number of distinct cloud regions you deploy workloads into across all providers. Each region adds routing complexity, data replication requirements, and latency considerations.

Cloud Accounts / Subscriptions ?Total number of cloud accounts, subscriptions, or projects across all providers. More accounts mean more IAM boundaries, billing entities, and networking silos to manage.

Cloud VPCs / VNets ?Total number of Virtual Private Clouds (AWS), Virtual Networks (Azure), or VPC Networks (GCP). Each is an isolated network requiring explicit peering or transit to communicate.

On-Prem VLANs ?Number of on-premises VLANs or network segments. Combined with cloud VPCs, this represents your total network segmentation surface.

Cross-Cloud Dependencies ?Services or applications that depend on resources in a different cloud or domain. For example, a frontend in AWS calling an API in Azure counts as one cross-cloud dependency.

Hybrid Connectors

hub

Connector Types: Cloud-to-datacenter connectors carry additional weight in the complexity model because they bridge fundamentally different control planes, requiring more operational coordination.

Cloud ↔ Cloud ?Direct connections between cloud providers — e.g., AWS Direct Connect to Azure ExpressRoute, or VPN tunnels between clouds. Includes transit gateway peering.

Cloud ↔ Data Center ?Connections between any cloud provider and your private DCs or colocations — e.g., Direct Connect, ExpressRoute, Cloud Interconnect to an on-prem network.

DC ↔ DC ?Direct connections between your data centers and/or colocation facilities — e.g., dark fiber, MPLS circuits, or SD-WAN tunnels between physical sites.

Governance & compliance

Security policies, compliance frameworks, and sovereignty requirements.

Security Policies ?Total number of managed security rules across your environment — firewall rules, security group rules, network ACLs, WAF rules, etc. A typical mid-size enterprise has 200–500.

Physical Compliance Requirements ?Number of distinct physical security compliance requirements you must meet — SOC 2 physical controls, ISO 27001 Annex A.11, NIST PE controls, TIA-942 DC tier ratings, etc.

Compliance Frameworks ?Regulatory and industry compliance frameworks your infrastructure must satisfy. Each framework multiplies governance overhead across every infrastructure domain.

SOC 2

ISO 27001

HIPAA

PCI DSS

GDPR

FedRAMP

DORA

NIST CSF

Sovereignty-Aware Routing ?Does your network enforce data residency rules — ensuring traffic stays within required jurisdictions? Critical for GDPR, DORA, and government workloads.

Fully Implemented

Partial

Not Implemented

Cross-Cloud Failover Tested ?Have you validated that workloads can automatically fail over between clouds or domains? Untested failover is the leading cause of extended outages in multi-cloud environments.

Tested & Validated

Partially Tested

Not Tested

Toolchain & operations

Networking tools, operational boundaries, and team structure.

Cloud Networking Tools ?Count of distinct cloud networking tools you manage — VPC consoles, transit gateways, SD-WAN controllers, load balancers, DNS services, CDNs, service meshes, etc. Each adds operational overhead that compounds non-linearly.

On-Prem Networking Tools ?Count of on-premises networking tools — physical routers, switches, firewalls, IPAM, DCIM, SD-WAN appliances, etc. Set to 0 if you have no on-prem infrastructure.

Operational Boundaries (Teams) ?Number of distinct teams responsible for network infrastructure — e.g., cloud team, on-prem team, security team, platform team. More boundaries mean more coordination overhead during incidents and changes.

Change Management Processes ?Number of separate change approval workflows for infrastructure changes — e.g., one for cloud, one for DC, one for security. More processes slow down response time and increase MTTR.

Interaction Complexity ?How much cross-team coordination friction exists? Low (1–10): teams work independently with clear handoffs. Medium (10–25): regular coordination needed. High (25–50): heavy matrix organisation with frequent cross-team dependencies.

10

AI infrastructure & data

GPU deployments, inference endpoints, and data gravity — these drive the AI complexity amplifier.

memory

AI Amplifier: AI workloads don't just add complexity — they multiply it. GPU topology awareness, inference latency sensitivity, and data gravity create non-linear scaling effects on your overall complexity score.

Cloud GPUs ?Number of GPU instances running in cloud environments — includes training clusters, inference GPUs, and GPU-enabled VMs across all cloud providers.

On-Prem GPUs ?Number of GPUs in your data centers or colo facilities. On-prem GPUs carry additional complexity due to physical fabric management, cooling, and power considerations.

Which domains have GPUs? ?Select every infrastructure domain where GPU instances are deployed. Options are based on infrastructure you selected in Step 2. The more domains with GPUs, the higher the routing complexity.

Select your infrastructure domains in Step 2 first.

Inference Endpoints ?Number of model inference endpoints serving production traffic — e.g., SageMaker endpoints, Vertex AI endpoints, self-hosted model servers. Each is a latency-sensitive dependency.

Which domains host inference endpoints? ?Select every domain where model inference endpoints serve production traffic. Options are based on infrastructure you selected in Step 2. Greater spread increases data transfer costs and latency variance.

Select your infrastructure domains in Step 2 first.

Latency-Critical Paths ?Number of network paths with hard latency requirements (e.g., < 10ms). Includes GPU interconnects, real-time inference pipelines, trading systems, and live media streams.

Data Volume (TB) ?Total data volume in terabytes across all storage systems. Larger data volumes create "data gravity" that resists workload mobility and increases migration costs.

Distinct Data Stores ?Number of distinct database or storage systems — RDS instances, S3 buckets with unique datasets, on-prem databases, data warehouses, data lakes, etc.

Your EMCI Results

Review your final score, what band it falls into, and the drivers behind it. Download a personalised report and schedule time with experts for remediation.

Final score & breakdown

Five severity bands — LOW, MODERATE, HIGH, SEVERE, CRITICAL — each with defined thresholds mapping to operational risk.

Get your personalised report + remediation support

See exactly where your multi-cloud complexity stands — across infrastructure, tools, AI, and geopolitical risk.

Tata Communications processes your personal information as described in the Privacy Policy.

Get your personalised report + remediation support

Report Unlocked!