AWS re:Invent 2025 - Agentic AI Meets Cybersecurity: eSentire’s Atlas AI Powered by Snowflake & AWS
Source: Dev.to
🦄 Making great presentations more accessible.
Overview
AWS re:Invent 2025 – Agentic AI Meets Cybersecurity: eSentire’s Atlas AI Powered by Snowflake & AWS
In this session, Matt Marzillo (Snowflake) and Dustin Hillard (CTO, eSentire) discuss integrating Snowflake’s AI capabilities with AWS for cybersecurity. Matt introduces Snowflake Cortex, which provides batch processing of unstructured data and agentic systems via Cortex Search, Cortex Analyst, and Cortex Agent. Dustin explains how eSentire uses Snowflake to consolidate security telemetry—processing ~20 TB daily from network, endpoint, and log sources. Their agentic system performs up to 30 tool calls per investigation, achieving 95 % alignment with senior analyst decisions while cutting investigation time from 10–15 minutes to seconds. This enables eSentire to expand into new markets (e.g., India, Saudi Arabia) through platform licensing and to leverage Snowflake Intelligence for internal analytics across Salesforce, ServiceNow, and Gong data.
This article is auto‑generated; minor typos or inaccuracies may be present.
Main Part
Snowflake’s AI Capabilities and AWS Integration: From Cortex to Agentic Systems
Hello, everyone. I’m Matt Marzillo, Principal Partner Engineer at Snowflake, joined by Dustin, a Snowflake and AWS customer with eSentire. I’ll give a brief overview of Snowflake’s AI offerings and how customers are integrating them with AWS. Dustin will follow with a detailed use case.
Snowflake’s Evolution
Snowflake began as a cloud‑native enterprise data warehouse, offering the scalability and low‑admin overhead of the cloud while preserving the familiar SQL experience. Over the past decade we’ve expanded to support Data Lakehouse, Iceberg, data mesh, application development, data sharing, collaboration, and traditional machine learning. Our AI story is now central.
Tight Integration with AWS
Snowflake runs on AWS infrastructure (S3, EC2) and integrates natively with many AWS services: S3, Kinesis, Data Firehose, SageMaker, Bedrock, Glue, and Glue Catalog. This deep partnership enables seamless data movement and AI workflow orchestration.
Cortex: The AI Layer
All AI functionality in Snowflake is branded Cortex. Rather than separate services, Cortex provides a unified experience for building data‑centric AI systems:
| Capability | Typical Use |
|---|---|
| Batch processing of unstructured data (LLM‑driven SQL, document processing, image labeling, audio transcription, translation) | Initial AI production workloads |
| Agentic systems (Cortex Search, Cortex Analyst, Cortex Agent) | Context‑aware, interactive AI assistants built on structured and unstructured data |
- Cortex Search – Hybrid search across unstructured content.
- Cortex Analyst – Generates highly accurate answers by grounding LLM responses in structured data.
- Cortex Agent – Orchestrates multiple tools, searches, and analysts to produce a final response, which can be materialized anywhere.
Snowflake Intelligence offers a native UI to interact with these agents directly within Snowflake, and the same agents can be accessed from Slack, Streamlit, or custom applications.
Extending Agents with Agentic Core
Customers are now extending Cortex agents into broader Agentic Core workflows, connecting Snowflake‑hosted AI systems with external orchestration platforms. Snowflake provides guides and quickstarts for building these integrated pipelines, enabling a seamless bridge between data‑centric AI inside Snowflake and larger agentic ecosystems.