Sudhanshu Sharma — Software Engineer

Engineering trust in AI systems by building security controls that enable organizations to adopt Generative AI safely and at scale. I specialize in AI Security, with hands-on experience designing and developing runtime guardrails that inspect prompts, responses, tool usage, and agent interactions across enterprise AI applications. My work focuses on leveraging Small Language Models (SLMs) for security use cases, including content inspection, policy enforcement, risk detection, and AI governance. I have extensive experience training, evaluating, and optimizing models to meet demanding accuracy, latency, and operational reliability requirements in production environments. My expertise spans the complete AI lifecycle—from model development and evaluation to deployment validation and production readiness. Working closely with engineering teams, I focus on ensuring AI systems perform reliably, securely, and at scale in real-world enterprise environments. I have extensive experience validating cloud-native AI platforms running on Kubernetes and AWS EKS, including distributed inference architectures built on Ray Serve. My work involves designing comprehensive test strategies, evaluating model accuracy and performance, validating autoscaling behavior, request routing, workload isolation, resiliency, and end-to-end system reliability. Beyond engineering, I actively engage with customers, security leaders, and technical stakeholders to demonstrate AI security capabilities, discuss real-world deployment strategies, and help organizations navigate the evolving AI risk landscape. I regularly translate complex technical concepts into business outcomes, provide guidance on AI adoption challenges, gather customer feedback, and contribute to shaping product strategy and roadmap discussions. This combination of deep technical expertise and customer engagement enables me to bridge the gap between innovation, security, and business needs. Prior to focusing on AI Security, I worked on enterprise Identity and Access Management solutions, contributing to Authentication, Authorization, and Accounting (AAA) platforms used by large global organizations. I am passionate about advancing AI Security through practical innovation, scalable architectures, and security-first design principles. My areas of interest include AI Security, Agentic AI Security, AI Guardrails, LLM/SLM Systems, Model Serving & Inference, Distributed AI Platforms, Cloud Security, Identity & Access Management, Security Architecture, and Trustworthy AI.

Stackforce AI infers this person is a SaaS AI Security expert with extensive experience in software testing and automation.

Location: Bengaluru, Karnataka, India

Experience: 12 yrs 1 mo