Normalized claim
Answer quality improvement: 20% increase
increased Companion's answer quality by 20 percent
Trend Micro builds an agentic AI security assistant with Amazon Bedrock and Amazon Neptune
Use case typeDocument reviewUpdated Jun 13, 2026
Trend Micro enhanced its Companion AI assistant in its Vision One security software with long-term and short-term agentic memory capabilities. The company used Amazon Neptune as a knowledge graph and experience layer to connect threat intelligence, alerts, investigations, and successful remediation patterns so analysts can get more explainable and personalized security recommendations. Trend Micro combined Amazon Bedrock, Amazon Textract, and Amazon Managed Streaming for Apache Kafka to orchestrate agentic workflows, clean and enrich user conversations, and feed metadata into the graph for long-term memory and correlation analysis.
- Organization
- Trend Micro
- Industry
- Tech & Comms
- Location
- United States
- Published
- May 2026
Reported outcomes
+20%
answer quality improvementQuality & accuracy
Strategic outcomes
New product / capabilityBuilt long-term and short-term agentic memoryCustomer experience & trustDelivered more explainable recommendationsSpeed & agilityEnabled faster access to prior casesInnovation & cultureShifted security operations proactive
Why do we believe this?Outcome claims, sources, and evidence checks
Why do we believe this deployment?Customer identity, provider attribution, maturity, and source checks
- Customer
- Trend Micro
- Provider
- AWS
- Maturity
- Unknown
- Linked source
- AWS Customer Stories
No explicit deployment-stage evidence found.
Customer identity supportedSource describes one deploymentMaturity evidence evaluated
Primary read
Use case focus
Showing 3 of 5
- 1Security Assistant
- 2Agentic Workflow
- 3GraphRAG
- Enhance a cybersecurity assistant with durable memory so it can provide more accurate, explainable, and personalized recommendations.
- Turn years of threat intelligence and investigation experience into reusable organizational knowledge.
- Help analysts detect and respond to threats more proactively across teams with varied skill levels.
- Used Amazon Neptune to store relationships across threat intelligence, alerts, devices, and investigation outcomes.
- Applied Amazon Bedrock to orchestrate agentic AI workflows and route user intents to backend agents.
- Used Amazon Textract to scan user conversations and remove personally identifiable information.
- Used Amazon Managed Streaming for Apache Kafka to extract metadata from log collectors and feed the graph.
- Implemented GraphRAG with Claude models through Amazon Bedrock and continuously enriched the graph from new conversations.
- Trend Micro increased Companion's answer quality by 20%.
- The system delivers more accurate, explainable, and personalized recommendations.
- Teams can access relevant prior cases and proven remediation approaches more quickly.
- Operations shift from reactive to proactive by democratizing security expertise.
Architecture
Trend Micro built a graph-driven memory architecture in which Amazon Neptune stores connected security knowledge and interaction history. Amazon Bedrock orchestrates agentic workflows and routes prompts to the appropriate backend service agents. Amazon Textract is used to scan conversations and remove PII, while Amazon MSK extracts metadata from log collectors that is stored in Neptune for long-term memory and correlation. The company implements GraphRAG with Claude models via Bedrock, and new conversations continuously enrich the knowledge graph.
Sources & evidence1
Evidence: Medium50/100Evidence strength
- Customer explicitly identified
- Primary source available
- Quantified outcome available
- Technical implementation details available
Type: Customer StoryPublished: May 27, 2026Publisher: AWSEvidence: PrimaryConfidence: High
AI-generated summary. Verify important details with the linked sources before relying on this case.
Explore related AI use cases
Customer evidence
Provider evidence
Browse the catalog
Was this useful?
Community
Comments
Loading comments...