Amazon Quick now integrates with Snowflake Cortex AI
Query Snowflake data and documents in plain language and automate compliance workflows—all from one AI workspace, in minutes.
View original announcement →Visual Summary
graph TD
A{{Amazon Quick + Snowflake Cortex AI}}:::announced
B((Enterprise Users)):::external
C([MCP Server OAuth]):::feature
D(Snowflake Cortex Analyst):::compute
E(Snowflake Cortex Search):::compute
F([Quick Flows]):::feature
G([Quick Chat]):::feature
H(Snowflake AI Data Cloud):::storage
I([Quick Spaces]):::feature
B ==>|"natural language"| A
A -->|"authenticates"| C
C ==>|"connects"| D
C ==>|"connects"| E
D -->|"queries structured"| H
E -->|"searches unstructured"| H
A -->|"orchestrates"| F
F -->|"Cortex Agents"| D
A -.->|"ad-hoc queries"| G
A -.->|"blends context"| I
classDef announced fill:#ff9900,stroke:#ec7211,color:#fff,font-weight:bold
classDef compute fill:#e3f2fd,stroke:#1565c0,color:#1565c0
classDef storage fill:#e8f5e9,stroke:#2e7d32,color:#2e7d32
classDef feature fill:#fff3e0,stroke:#e65100,color:#e65100
classDef external fill:#f5f5f5,stroke:#616161,color:#616161
What's New
Amazon Quick now integrates with Snowflake Cortex AI via the Model Context Protocol (MCP), allowing teams to query structured Snowflake data and unstructured documents using natural language directly within their Quick workspace. The integration enables users to build automated, multi-step Flows that orchestrate Snowflake Cortex Agents, while also supporting ad-hoc conversational queries through Quick Chat. Authentication is handled via OAuth using Snowflake's managed MCP server, requiring no custom connectors.
How It Works
- MCP Connection Setup: Users configure a connection to Snowflake's managed MCP server using OAuth authentication, establishing a secure, standards-based channel between Amazon Quick and Snowflake Cortex AI.
- Structured Data Queries via Cortex Analyst: Once connected, Quick routes natural language questions about structured data to Snowflake Cortex Analyst, which translates them into SQL and returns tabular results without requiring users to write queries manually.
- Unstructured Document Retrieval via Cortex Search: For document-based questions, Quick routes prompts to Snowflake Cortex Search, which performs semantic search over unstructured content stored in Snowflake and returns relevant passages or summaries.
- Automated Flows with Cortex Agents: Users can build Quick Flows that orchestrate Snowflake Cortex Agents to execute repeatable, multi-step workflows—spanning both structured and unstructured data—with consistent, governed output.
- Intelligent Prompt Routing: Quick automatically determines which prompts should be directed to Snowflake Cortex AI versus other connected sources, blending Snowflake results with enterprise knowledge stored in Quick Spaces for contextualized answers.
- Conversational Access via Quick Chat: The same MCP connection is available in Quick Chat, enabling ad-hoc follow-up questions and exploratory data conversations alongside automated flows without any additional configuration.
Why It's Important
- Eliminates Manual Data Bridging: Teams that currently switch between Snowflake dashboards and document repositories to answer business questions can now do both conversationally in a single interface, reducing context-switching and manual effort.
- Accelerates High-Stakes Workflows: The AML alert triage use case demonstrates the real-world impact—automated workflows reduced investigation time from 30–90 minutes to under 5 minutes, a transformational efficiency gain for compliance-heavy industries.
- Governs AI Workflows at Scale: By building Flows that orchestrate Cortex Agents with structured output, organizations gain repeatable, auditable AI processes rather than one-off ad-hoc queries, which is critical for regulated industries.
- Leverages Existing Snowflake Investments: Organizations already running data workloads on Snowflake can extend their AI Data Cloud capabilities into Quick without migrating data or rebuilding pipelines, protecting existing investments.
- Unifies Structured and Unstructured Intelligence: Most enterprise decisions require both quantitative data and qualitative documents; this integration is one of the few that natively bridges both in a single governed workflow.
How It's Different
- Standards-Based Protocol (MCP): Unlike proprietary connector approaches, this integration uses the open Model Context Protocol standard, making it more portable and reducing vendor lock-in compared to custom API integrations.
- No Custom Connectors Required: Quick Flows translates user requests into standardized MCP calls automatically, eliminating the engineering overhead typically associated with building and maintaining point-to-point integrations.
- Dual-Mode Access (Flows + Chat): The same single MCP connection powers both structured automated workflows and free-form conversational queries, whereas most integrations serve only one interaction mode.
- Context Fusion with Quick Spaces: Unlike a direct Snowflake query tool, Quick blends Snowflake Cortex results with enterprise knowledge indexed in Quick Spaces, delivering answers that combine data warehouse insights with broader organizational context.
- Intelligent Routing Without User Configuration: Quick automatically determines when to invoke Snowflake Cortex AI versus other sources, sparing users from manually selecting tools or writing routing logic themselves.
- Managed OAuth Security: Authentication is handled through Snowflake's managed MCP server with OAuth, providing enterprise-grade security without requiring users to manage API keys or custom credential stores.
When to Prefer It
- AML and Compliance Investigations: Financial institutions triaging anti-money laundering alerts benefit immediately, as the workflow spans structured transaction data (Cortex Analyst) and unstructured case notes or regulatory documents (Cortex Search).
- FinOps Cost Triage: Teams analyzing cloud or operational costs stored in Snowflake alongside unstructured vendor contracts or policy documents can automate investigation and reporting in a single Flow.
- SRE Incident Response: Site reliability engineers who need to correlate structured metrics and logs in Snowflake with unstructured runbooks or post-mortems can build governed, repeatable triage workflows.
- Recurring Cross-Data Reporting: Any team that runs the same multi-step report combining Snowflake data with document-based context on a regular cadence should build a Quick Flow to eliminate manual repetition.
- Ad-Hoc Snowflake Exploration Without SQL Skills: Business users who need to explore Snowflake data conversationally without writing SQL can use Quick Chat with the MCP connection as a natural language query interface.
- Enterprises Already on Both AWS and Snowflake: Organizations with existing footprints on both platforms can activate this integration with minimal setup, making it the lowest-friction path to AI-powered cross-data workflows.
Availability
- GA Status: Generally available as of June 11, 2026; no preview or beta designation is indicated in the announcement.
- Regional Availability: Available in all AWS Regions where Amazon Quick is supported, including US East (N. Virginia), US West (Oregon), and multiple Asia Pacific and European regions; note that some regions support only QuickSight features within the Quick platform.
- Pricing: No separate integration fee is announced; access is included as part of Amazon Quick, which offers a free trial; Snowflake Cortex AI usage costs are governed by Snowflake's own pricing.
- Authentication Requirement: Requires OAuth setup via Snowflake's managed MCP server; users must have an active Snowflake account with appropriate Cortex AI entitlements enabled.
- Limitation – Snowflake Dependency: The integration relies on Snowflake Cortex AI capabilities (Cortex Analyst, Cortex Search, Cortex Agents), so organizations without a Snowflake environment cannot use this integration.
