Cloud provider comparison
Across 3,983 eligible AI deployments, Microsoft has the highest eligible deployments at 2177.
Compare source-linked enterprise AI deployments across Microsoft, AWS, Google Cloud, Oracle, IBM, Alibaba Cloud. Review complete evidence coverage alongside sampled technology, capability, and implementation patterns.
3,983
784
1,436
2,294
Source-linked enterprise AI deployments using Microsoft Azure, Azure AI Foundry, Azure OpenAI, Copilot, Microsoft Fabric, and related services.
2,177
Eligible cloud provider deployments
Sampled technology signals
Source-linked enterprise AI deployments using Amazon Web Services, including Bedrock, SageMaker, Amazon Q, and cloud-native AI services.
837
Eligible cloud provider deployments
Sampled technology signals
Source-linked enterprise AI deployments using Google Cloud, including Gemini, Vertex AI, BigQuery, and enterprise AI services.
700
Eligible cloud provider deployments
Sampled technology signals
Source-linked enterprise AI deployments using Oracle Cloud Infrastructure, including OCI Generative AI, Select AI, Oracle Digital Assistant, and AI embedded in Fusion Applications.
27
Eligible cloud provider deployments
Sampled technology signals
Source-linked enterprise AI deployments using IBM watsonx, including watsonx.ai, watsonx Orchestrate, watsonx Assistant, and Granite models.
50
Eligible cloud provider deployments
Sampled technology signals
Source-linked enterprise AI deployments using Alibaba Cloud, including Model Studio, Qwen (Tongyi Qianwen) models, and Platform for AI (PAI).
192
Eligible cloud provider deployments
Sampled technology signals
Cases by provider · segmented by top technologies
Colour marks the provider; segments run largest-first and fade with rank, so the most saturated one leads each provider. Hover or focus a segment for its count and share.
Technology readout
Microsoft has the highest eligible deployments at 2177. This uses the complete eligible cloud provider dataset.
Alibaba Cloud spans 196 sampled technologies, led by Elastic Compute Service, Object Storage Service.
Alibaba Cloud has the highest recent share at 67.2%. This uses the complete eligible dataset within the configured recent window.
Microsoft shows the broadest sampled technology adoption context across 16 industries and 24 countries.
Side-by-side table
| Provider | Eligible deployments | Share | Recent deployments | Recent share | Quantified outcomes | Production evidence | Sampled innovation | Top technologies | Technology breadth | Capability pattern | Top industry |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Microsoft | 2,177 | 55% | 101 | 5% | 587 | 1,504 | Not scored | Azure OpenAI, Copilot Studio, Microsoft 365 Copilot | 159 | Agents, Copilot | Professional Services |
| 837 | 21% | 299 | 36% | 397 | 352 | Not scored | Amazon Bedrock, Amazon S3, AWS Lambda | 170 | Agents, RAG | Tech & Comms | |
| GCP | 700 | 18% | 236 | 34% | 371 | 339 | Not scored | Gemini Enterprise Agent Platform, BigQuery, Gemini | 182 | Agents, Multi-agent | Tech & Comms |
| 27 | 1% | 7 | 26% | 4 | 2 | Not scored | Oracle Digital Assistant, Oracle Fusion Cloud HCM, OCI Generative AI | 70 | Agents, RAG | Healthcare | |
| 50 | 1% | 12 | 24% | 15 | 9 | Not scored | IBM watsonx Orchestrate, IBM watsonx.ai, IBM Cloud | 64 | Agents, Multi-agent | Finance | |
| Alibaba | 192 | 5% | 129 | 67% | 62 | 88 | Not scored | Elastic Compute Service, Object Storage Service, Alibaba Cloud Elastic Compute Service | 196 | Vision, RAG | Finance |
Crawlable comparison pages
Evidence examples
Rivian deployed an AI-driven ERP agent to automate purchase-order accrual accounting for custom tooling, a finance workflow spanning many active purchase orders and long supplier delivery cycles. The agent interprets finance standard operating procedures, handles SAP exceptions, gathers purchase-ord
Canva built generative AI tools on Amazon Bedrock to help creators move from a blank page to finished designs by generating text, images, and recommendations.
Variational AI developed Enki, a generative AI platform for designing novel small-molecule candidates based on target properties. Enki's foundation model was trained on published, approved, and patented drug data and generates candidates optimized across efficacy, safety, potency, and selectivity. F
Voxelis AI developed VoxVision, an aircraft-mounted system that analyzes thermal imagery and environmental data to provide wildfire-response intelligence. Data captured aboard aircraft and AI-generated insights are processed and shared through AWS infrastructure with authorized response personnel.
Waabi uses AWS infrastructure to run simulation, testing, and training workflows for its autonomous-driving platform. Waabi World enables the company to validate safety-critical driving scenarios in simulation before real-world deployment.
Nestlé tested GPU-accelerated watsonx.data against its global order-to-cash data mart, which covers orders, fulfillment, deliveries and invoices across 186 countries. The workload-specific test evaluated GPU acceleration for refreshing Nestlé's global analytics data mart.
Project Courage, a Connecticut addiction and mental health center, consolidated disconnected paper-based and digital workflows into a shared digital platform for behavioral health operations. The platform provided a common foundation for clinical intake, treatment episodes and documentation, and fin
Unity China leverages Alibaba Cloud Model Studio to deploy Qwen and third-party foundation models for Tuanjie AI, its AI-native engineering platform. Tuanjie AI automates workflows including code comprehension, real-time editing, and debugging, while supporting collaboration across development teams
Onc.AI develops imaging biomarkers to help oncologists and pharmaceutical researchers predict patient survival and treatment response. The company deployed its enterprise MLOps platform on Oracle Cloud Infrastructure and moved workloads to OCI GPU instances for model training and clinical deployment
FAQ
Microsoft has the highest eligible deployments at 2177.
Alibaba Cloud has the highest recent share at 67.2%.
The page compares complete eligible deployment totals, recent share, quantified-outcome coverage, and explicit production evidence for all six tracked providers. Technology, industry, country, partner, and innovativeness signals are calculated from representative source-linked records and labelled as sampled.
Use the evidence links on each provider card or open Deployments with the same provider and industry filters applied.
