Microsoft unveiled Scout, a new enterprise AI assistant, and RTX Spark, a hardware platform for AI workloads, at its Build 2026 developer conference in Seattle this week. The announcements position Microsoft to compete more directly with Google’s Gemini for Workspace and Amazon’s Q for Business in the enterprise AI assistant market.
Scout integrates across Microsoft’s developer tools and enterprise applications, according to demonstrations shown at the conference. The assistant can generate code, analyse business data, and automate workflows within Azure, Visual Studio, and Microsoft 365 environments. Unlike Copilot, which targets individual productivity, Scout is designed for team-based development and cross-functional business processes.
The RTX Spark hardware represents Microsoft’s first purpose-built AI inference chip for edge deployment. The company stated that RTX Spark delivers 40 teraflops of AI performance whilst consuming under 75 watts, making it suitable for on-premises enterprise deployments where cloud connectivity is limited or data sovereignty requirements prevent cloud processing.
Microsoft’s timing reflects intensifying competition in enterprise AI infrastructure. Google announced expanded Gemini enterprise features in March 2026, whilst Amazon has been aggressively pricing Q for Business to capture market share. Salesforce’s Einstein GPT continues to gain traction in customer relationship management workflows, creating pressure on Microsoft to differentiate its AI offerings beyond the consumer-focused Copilot brand.
The business impact centres on three constituencies. Enterprise customers gain more deployment flexibility, particularly in regulated industries like healthcare and finance where on-premises AI processing addresses compliance requirements. Microsoft partners, especially systems integrators and independent software vendors, acquire new implementation and customisation opportunities around Scout’s APIs. Competitors face renewed pressure as Microsoft leverages its existing enterprise relationships and Azure infrastructure to bundle AI capabilities.
Developer adoption will prove critical. Microsoft announced that Scout’s API will be available in public preview in July 2026, with general availability planned for Q4 2026. The company is offering RTX Spark development kits to enterprise customers with existing Azure commitments above $100,000 annually, a threshold that encompasses approximately 15,000 organisations globally based on Microsoft’s fiscal 2025 Azure revenue disclosures.
Technical specifications reveal Scout’s architecture relies on a mixture-of-experts model approach, routing queries to specialised sub-models based on task type. This differs from the monolithic model approach used by competitors and theoretically provides cost advantages for enterprises running diverse workloads. RTX Spark supports both Microsoft’s proprietary models and open-source alternatives including Llama 3 and Mistral, addressing customer concerns about vendor lock-in.
The hardware strategy marks a notable shift. Microsoft has historically relied on Nvidia, AMD, and other chip manufacturers for AI acceleration. RTX Spark, whilst not replacing these partnerships, gives Microsoft greater control over the edge computing stack and reduces dependency on external suppliers for specific use cases. Industry observers note this follows similar moves by Google with its Tensor Processing Units and Amazon with its Trainium chips.
Pricing remains undisclosed for both products. Microsoft indicated that Scout will follow a consumption-based model similar to Azure OpenAI Service, whilst RTX Spark will be available through both purchase and lease arrangements. The company emphasised that existing Copilot for Microsoft 365 licences will not automatically include Scout access, positioning it as a distinct enterprise offering.
The market will be watching three indicators: enterprise adoption rates for Scout compared to Copilot, which has seen mixed uptake since its 2023 launch; competitive responses from Google and Amazon, particularly around pricing and feature parity; and whether RTX Spark gains meaningful deployment beyond Microsoft’s immediate customer base. Developer feedback during the July preview period will signal whether Scout’s API design meets enterprise integration requirements or requires substantial revision before general availability.
