# Tamnoon builds AI remediation agent Tami on Amazon Bedrock to cut cloud risks by 90%

Digest AI · Enterprise & Industry · published 2026-10-14T13:53:33Z

Canonical: https://digestai.news/story/tamnoon-builds-ai-remediation-agent-tami-on-amazon-bedrock-to-cut-clou

## Summary

Tamnoon, an AI-powered cloud security platform, introduced **Tami**, an agentic remediation system built on **Amazon Bedrock** that autonomously fixes critical cloud misconfigurations. The system processes findings from tools like **Wiz, CrowdStrike, and AWS Security Hub**, then investigates, scores safety, and executes fixes—all without touching production—using specialized AI agents for each step. Tamnoon claims Tami reduced critical cloud exposures by **over 90%** in **90 days** for customers, with zero production incidents, by combining **LLM-as-a-judge validation**, guardrails, and human-in-the-loop oversight for high-risk cases.

The architecture leverages **Amazon Bedrock’s multi-model API** to match tasks to the best foundation model, from deep investigative reasoning (using **Anthropic’s Claude**) to high-volume validation and code review. Security is hardened through **IAM-scoped permissions**, **Bedrock Guardrails**, and **cross-region inference** to ensure data stays within AWS boundaries. Tamnoon’s **Remediation Confidence Indicator (RCI)**—**SAFE, RISKY, or AWAITING DATA**—guides automation, while **backup, verify, and rollback** steps mitigate risks. The platform processes **14 million+ alerts** across customers, with results typically visible within **90 days**, according to Tamnoon’s internal research.

## Key points

- Tami cuts critical cloud exposures by over 90% in 90 days, per Tamnoon’s data, with zero production issues
- Uses Amazon Bedrock’s multi-model API to assign tasks to specialized AI agents (e.g., investigation, safety scoring, code review)
- Security features include IAM permissions, Bedrock Guardrails, and tenant-isolated AWS environments

## Why it matters

Tamnoon’s approach demonstrates how AI agents can safely automate cloud security remediation at scale, reducing manual triage and risk accumulation. For enterprises, it validates that frontier AI—when paired with strict governance—can close real-world security gaps without downtime.

## Sources

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 Machine Learning Blog, 2026-10-14, primary source)

Part of the developing story: [AWS Expands Bedrock AgentCore Ecosystem](https://digestai.news/thread/aws-shows-synthetic-monitoring-with-nova-act-and-agentcore) (6 stories)

## Cite

Digest AI, "Tamnoon builds AI remediation agent Tami on Amazon Bedrock to cut cloud risks by 90%", 14 October 2026, https://digestai.news/story/tamnoon-builds-ai-remediation-agent-tami-on-amazon-bedrock-to-cut-clou

---

Written by Digest AI's editorial model from the linked sources; the sources are the record. Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse
JSON: https://digestai.news/story/tamnoon-builds-ai-remediation-agent-tami-on-amazon-bedrock-to-cut-clou.json