How Greenlight Financial built an autonomous SOC solution on Amazon Bedrock AgentCore
Greenlight Financial resolves more than 90% of up to 4,000 monthly security and risk cases without human intervention using Luna, a tiered multi-agent system on Amazon Bedrock AgentCore. This post covers the architecture, episodic memory for continuous learning, and the expansion from SOC operations into financial risk triage.
First seen by Oriel on .
AI summary
Greenlight Financial built Luna, a tiered multi-agent SOC on Amazon Bedrock AgentCore that resolves more than 90 percent of up to 4,000 monthly security and risk cases autonomously.
- Luna’s Tier 1 supervisor orchestrates more than 30 specialized agents, while Tier 2 investigates low-confidence cases across three to five data sources.
- AgentCore runtime hosts the agent code in private VPC subnets, while AgentCore memory supports episodic and long-term learning, deduplication, and resolution templates.
- A Lambda, DynamoDB, and feedback pipeline increased autonomous resolution from 83 percent with Tier 1 to more than 90 percent with Tier 2 and learnings.
- Routine triage decreased from 15–30 minutes to seconds, avoiding an estimated 900–1,800 analyst hours monthly across 3,000–4,000 cases.
- Luna recommends and supports remediation rather than acting autonomously without policy approval, preserving human oversight for regulated security and financial-risk workflows.
#amazon bedrock #multi-agent systems #security operations #financial risk
Generated by gpt-5.6-sol. Check the original for details.
Aliases
| Source | External ID | URL |
|---|---|---|
| AWS Machine Learning Blog | c5e92a0371483b3162fb07723de361bea61a1cce | https://aws.amazon.com/blogs/machine-learning/how-greenlight-financial-built-an-autonomous-soc-solution-on-amazon-bedrock-agentcore/ Read discussion |