Technology Radar

Microsoft's cross-platform development platform for backend services, web applications and cloud-native workloads.

The data foundation AI depends on: quality, lineage, permission metadata, freshness and semantic structure.

The broadest public cloud platform, spanning compute, storage, data, machine learning and managed services.

A distributed event streaming platform used as the backbone for real-time data movement and event-driven architecture.

Architectural patterns for decomposing systems into independently deployable services with governed interfaces.

Semantic models, reporting and visualization that turn governed data into decisions people actually make.

Containerized packaging and the orchestration platform that has become the default substrate for cloud-native workloads.

Ownership, definitions, quality, lineage and access control: the practice that makes a data estate usable and defensible.

A lakehouse platform unifying data engineering, analytics and machine learning on open storage formats.

A transformation framework that brings software engineering practice — version control, testing, documentation — to analytics SQL.

Large language and multimodal models applied to enterprise knowledge work, with retrieval, evaluation and human oversight.

A mature, statically typed language and runtime underpinning a large share of enterprise backend systems.

The language of the web, with TypeScript adding static typing for large, long-lived codebases.

Microsoft's public cloud, closely integrated with the Microsoft enterprise estate, identity and data platform.

Microsoft's business application suite spanning ERP and CRM, tightly integrated with Power Platform, Microsoft 365 and Azure.

An integration and API management platform for connecting applications, data and devices across hybrid estates.

A JavaScript runtime for server-side applications, well suited to I/O-bound API and integration workloads.

Enterprise applications and database technology spanning ERP, supply chain, HCM, customer experience and enterprise performance management.

Classical machine learning for forecasting, classification, anomaly detection and optimization: the least fashionable AI, and often the highest return.

The dominant language for data engineering, machine learning and AI, also widely used for backend services and automation.

The component-based UI library and the application framework built on it, with server rendering and hybrid data fetching.

The policies, controls, testing and oversight that make enterprise AI adoption defensible and auditable.

Grounding language model output in retrieved enterprise content, with citations and permission-aware access.

The customer relationship management platform for sales, service, marketing and customer-facing digital experiences.

The enterprise resource planning suite that runs finance, supply chain and operations for a large share of global manufacturing and process industry.

A cloud data platform separating storage from compute, with strong data sharing and workload isolation.

A batteries-included front-end framework with strong conventions, favoured for large, long-lived enterprise applications.

Building systems from independently replaceable, API-first components rather than from monolithic suites.

Google's public cloud, with particular strength in data analytics, machine learning and container orchestration.

Coordinating process mining, workflow, RPA, document AI and agents as one automation portfolio rather than as separate tools.

Visual development platforms that shorten delivery for the long tail of internal business applications.

Microsoft's unified analytics platform bringing data engineering, warehousing, real-time analytics and BI onto one storage layer.

Low-code application, automation and analytics tooling sharing a common data platform with Dynamics 365 and Microsoft 365.

A workflow platform originally for IT service management, now used to digitize service and operational workflows across the enterprise.

An enterprise robotic process automation platform with process mining, document understanding and orchestration.

AI systems that plan and execute multi-step tasks using tools, with approval gates and observability.

A cloud-native robotic process automation platform with document processing and analytics.

Models of physical assets or processes kept synchronized with reality by operational data.

Running inference on or near the device, where latency, bandwidth, privacy or resilience rules out the cloud.

Models that process and relate multiple input types — text, image, audio, video — within a single system.

Techniques that allow data to be used for analysis and testing while reducing exposure of the underlying personal information.

Software, perception and fleet integration for systems that operate with limited human intervention.

Distributed ledger technology, applicable where multiple parties need shared records without a trusted intermediary.

Designing and operating software for lower energy and carbon intensity — which usually also means lower cost.

Preparing cryptographic estates for post-quantum algorithms, and assessing where quantum computing may eventually apply.

Augmented, virtual and mixed reality applied to training, remote assistance and complex visualization.