Ms. Anoosha Cherukuri

Secure Software Development Lifecycle (SSDLC) for Data Science Models

Abstract:

While traditional Secure Software Development Lifecycles (SSDLC) focus on securing code and infrastructure, they often fail to address the unique risks posed by data science models. This work proposes an extended SSDLC framework that integrates data-centric and model-centric security practices into each stage of the development lifecycle. By adapting principles from standards like OWASP SAMM, the framework introduces stage-gated security controls—such as secure data sourcing, model integrity validation, adversarial testing, and post-deployment drift detection—within modern DevSecOps pipelines. This approach bridges the existing gap between software engineering and machine learning operations, offering a structured methodology to build and deploy secure, trustworthy AI systems.  

Profile:

Anoosha Cherukuri is a Senior Software Developer with over a decade of experience in full-stack development and public sector digital modernization. She currently leads cloud-native platform initiatives at the State of Tennessee, where she integrates scalable architecture and AI-driven components to improve service delivery and system efficiency. Her expertise spans frontend and backend development, API design, and intelligent system integration.