CloudSentinel: An AI-Assisted Cloud Misconfiguration Detection and Remediation Framework for Resource-Constrained Engineering Teams

Cloud misconfiguration remains a persistent security challenge for small software development teams operating without dedicated security expertise. This paper presents CloudSentinel, an AI-assisted cloud misconfiguration detection and remediation framework that improves the actionability of security findings while preserving reliable detection and human control. CloudSentinel utilizes a deterministic Python and boto3 scanning engine to assess AWS configurations across S3, IAM, Security Groups, EBS, RDS, and CloudTrail, while a Large Language Model (Llama 3.3 70B via the Groq API) is constrained to generate plain-language explanations and remediation guidance downstream from confirmed findings. Evaluation against Prowler produced a 98.3% detection agreement rate across 60 directly comparable control-resource pairs. Furthermore, CloudSentinel achieved an average actionability score of 4.67 out of 5 across six representative findings, compared to 0.83 for Prowler’s raw output. The research also establishes four core operational safeguards for responsible AI-assisted security engineering: data minimisation, independent verification, evidence-paired explanation, and a strict human-in-the-loop automation boundary.