Maximizing Snowflake Collaboration via Teams and Email
- Snowflake has integrated its Cortex Agents with Microsoft Teams and Microsoft 365 Copilot, allowing business users and non-technical staff to interact with structured and unstructured data using natural...
- The integration is available for deployment via Microsoft AppSource.
- Cortex Agents function as a stateless REST API that orchestrates a combination of tools to deliver insights.
Snowflake has integrated its Cortex Agents with Microsoft Teams and Microsoft 365 Copilot, allowing business users and non-technical staff to interact with structured and unstructured data using natural language. This integration aims to reduce context-switching by bringing data insights directly into the communication platforms where business decisions are made.
The integration is available for deployment via Microsoft AppSource. It embeds Snowflake’s conversational AI agents into the Microsoft 365 ecosystem, including services such as Word, Excel, and Outlook, as well as Microsoft Teams chats.
Technical Architecture of Cortex Agents
Cortex Agents function as a stateless REST API that orchestrates a combination of tools to deliver insights. The system utilizes a hybrid approach to handle different types of data sources through two primary tools: Cortex Analyst and Cortex Search.
Cortex Analyst is used for structured data, where it generates SQL to process the information. Snowflake reports that this tool achieves 90%+ accuracy in SQL generation via semantic models. Cortex Search is employed for unstructured data, extracting relevant insights from non-tabular sources.
The agents manage the entire workflow from natural language to SQL conversion, context retrieval, and LLM orchestration. To ensure the quality of the responses, the system includes features such as in-line citations, multi-message conversation context management, and answer abstention for queries that are deemed irrelevant.
Impact on Business Workflows
The primary goal of this integration is to move beyond traditional BI dashboards and reports, which often require business users to wait for data analysts to provide updates. By providing a natural language interface, organizations can enable users to query data and receive direct answers and visualizations in near real time.
Potential applications for these agents include searching through sales conversations or blending structured and unstructured data to optimize supply chain operations and accelerate customer support.
Developer Implementation and Deployment
For developers, the integration is designed to reduce latency and complexity. The use of a single API call allows for real-time streamed responses and simplified connectivity configuration.
The setup process involves establishing the Snowflake environment and configuring app connectivity to link the Cortex Agent App with Microsoft Teams or Microsoft 365 Copilot.
Data Privacy and Security Boundaries
Snowflake has specified that when using this integration, data is sent and received between the Snowflake Service and Microsoft services.
According to Snowflake’s documentation:
Snowflake is not responsible for the privacy, security, or integrity of data once it leaves the Snowflake Service boundary. Your use of Microsoft Teams or Microsoft 365 Copilot, and any data you process with it, is governed solely by the terms between you and Microsoft.
Snowflake Documentation