Snowflakespecialist in AI Data Cloud, Announced Snowflake Cortex AI for Financial Servicesa comprehensive solution that brings together AI capabilities and strategic alliances so that companies can unify their financial data ecosystem and securely deploy AI models, applications and agents with that data. These functions have extensive security controls, in addition to meeting the standards required in regulated industries.
The company has also announced a new managed Model Context Protocol (MCP) server, which is now available in public testing. With it, organizations can leverage their proprietary and third-party data in Snowflake. Customers will be able to use this MCP server to connect their data to agent applications and platforms, such as Anthropic, CrewAI, Cursor, Salesforce Agentforce, and Windsurf, through the managed MSP server, to build high-context AI agents and applications.
Cortex AI for Finance enables business agents to accelerate complex financial tasks, such as market analysis, quantitative research, fraud detection, customer service, and claims management. This saves companies time, reduces operating costs and delivers information more quickly. Snowflake’s MCP Server expands this capability by enabling industry-wide interoperability, securely connecting to Snowflake data as well as third-party data and applications.
The Cortex AI for Financials ecosystem delivers high-quality, trusted data from both structured and unstructured financial data providers and publishers that organizations can integrate with their AI applications and agents.
By combining sector-specific data from leading financial institutions and publishers (such as market analysis, expert research, business content, and news) with their proprietary data in Snowflake, financial services companies can gain deeper insights, greater accuracy, and better results from their AI.
Cortex AI’s complex machine learning workflows for financial services will be made easier with Snowflake Data Science Agent, which acts as an AI coding agent, automating data cleansing, feature engineering, model prototyping, and validation.
This allows teams to move from raw data to production-ready models more quickly. This means automating and optimizing the models that underpin quantitative research, fraud detection, 360 customer insights, and underwriting workflows.
When it comes to unstructured data analysis, Snowflake Cortex AISQL, which adds features such as AI-powered extraction and transcription, allows users to process and extract information from documents, audio and images efficiently and at scale. In this way, workflows can be transformed from start to finish. Such as customer service, investment analysis, claims management and choosing the next best action.
While Data Science Agent and Cortex AISQL accelerate workflows for technical and research teams, Snowflake Intelligence, in public testing, offers business users an intuitive conversational interface to obtain natural language insights from data stored in Snowflake. Also third-party data, applications and agents.
This allows users to quickly discover actionable information from both structured tables and unstructured documents. In this way, access to data and information is democratized across all financial institutions, and the technical overload that slows down business decision-making is eliminated.
AI agents extend the capabilities of LLMs by interacting with external tools, completing complex workflows, and understanding the broader context of an organization. However, connecting these AI agents to existing enterprise systems has presented challenges. The MCP has emerged in recent months as a solution to this problem, offering a standardized way for LLMs to integrate with data, APIs, and services.
With the introduction of Snowflake’s MCP Server, businesses can enable the MCP Server to connect with tools built on top of Snowflake. Snowflake’s MCP Server connects Cortex Analyst and Cortex Search with external AI agents through a standards-based MCP interface, unifying the retrieval of structured and unstructured data. This simplifies enterprise application architecture and eliminates the need for custom integrations, accelerating the delivery of context-rich AI applications and agents.
It also allows access to proprietary and third-party data shared in Snowflake with external tools. With Snowflake’s MCP server, remote agents can connect to Snowflake data as well as third-party data shared from the Snowflake Marketplace through Cortex Knowledge Extensions, enabling interoperability with the broader AI ecosystem.
Snowflake MCP Server can be used to connect to various applications and agent platformsamong those that are Anthropic, Augment Code, Amazon Bedrock AgentCore, CrewAI, Cursor, Devin by Cognition, Glean, Kumo, Mistral, Agentforce by Salesforce, UiPath, Windsurf, Workday and WRITER.
