# Tamnoon uses Amazon Bedrock to cut cloud security alerts by 90% in three months

> Tamnoon’s AI agent Tami automates investigation and safe remediation of cloud‑security findings, delivering over a 90% drop in critical exposures within 90 days.

Oossa · 2026-10-14 · https://oossa.com/en/tamnoon-uses-amazon-bedrock-to-cut-cloud-security-alerts-by-90-in-three-months

Tamnoon announced that its AI‑driven remediation system, Tami, runs on Amazon Bedrock and has reduced critical cloud exposures by more than 90% in a 90‑day period. The platform pulls findings from CNAPP tools such as Wiz, CrowdStrike and GuardDuty, uses four specialized agents to investigate, score safety, plan fixes and generate code, and executes changes with built‑in guardrails. All inference runs inside the customer’s AWS account, keeping data inside the AWS trust boundary.

## The facts

- 90% reduction in critical cloud exposure in 90 days
- Built on Amazon Bedrock using Anthropic Claude via the ChatBedrockConverse API

## Why it matters

Security teams can remediate thousands of cloud misconfigurations automatically without risking production downtime.

## Sources & references

1. [How Tamnoon built AI-powered cloud security remediation with Amazon Bedrock](https://aws.amazon.com/blogs/machine-learning/how-tamnoon-built-ai-powered-cloud-security-remediation-with-amazon-bedrock/) – AWS, 2026-10-14

Last updated: 2026-10-14