Tamnoon has built Tami, an agentic cloud remediation orchestrator that runs on Amazon Bedrock, to close the gap between detection and remediation of cloud security findings. Tami autonomously ingests findings from CNAPPs and cloud‑native tools, enriches them with asset and exposure context, prioritizes them, and then generates and orchestrates fixes with guardrails and governance. The platform uses four specialized AI agents, each running in a serverless, event‑driven architecture on AWS, to handle distinct phases of the remediation pipeline. By leveraging Bedrock’s single API, cross‑region inference, and guardrails, Tamnoon can match each task to the most suitable foundation model without complex integrations, enabling easy model upgrades and efficient scaling. The solution achieved an over 90 percent reduction in critical cloud exposure in 90 days with zero production issues, demonstrating the effectiveness of automated, AI‑driven remediation at scale. Tamnoon’s architecture isolates the platform within its own AWS accounts, uses Amazon EKS for always‑on services, Lambda for event‑driven tasks, and a suite of AWS messaging and workflow services to process large volumes reliably. Security is maintained through IAM‑scoped Bedrock inference, guardrails that review requests and generated code, and a defense‑in‑depth safety architecture that separates read‑only investigation from write operations, ensuring that no remediation is applied without a prior safety assessment. The platform integrates with a range of detection tools and delivers outcomes to common collaboration and ticketing systems such as Slack, Jira, ServiceNow, and GitHub. This approach showcases how AI can be integrated into cloud security operations to automate complex remediation tasks while maintaining stringent security controls.
How Tamnoon built AI-powered cloud security remediation with Amazon Bedrock
Tamnoon has created Tami, an AI‑powered cloud remediation orchestrator built on Amazon Bedrock, to automatically investigate, prioritize, and safely remediate misconfigurations reported by CNAPPs and other security tools. By ingesting findings, enriching them with context, and delegating work to specialized agents, Tami achieved over a 90% reduction in critical cloud exposure within 90 days without any production incidents. The solution leverages Bedrock’s single‑API model access, cross‑region inference, and guardrails to maintain security and scalability.
Why it matters
This demonstrates how large‑language‑model platforms can be harnessed to automate complex security workflows at scale, reducing human workload and accelerating remediation in multi‑cloud environments.