Sharad Nagariya
Member for
10 months 1 weekSharad is a distinguished technology leader renowned for driving digital transformation and groundbreaking innovation, with a profound specialization in Artificial Intelligence. He is at the forefront of AI research, particularly in the development of advanced AI agent frameworks, leveraging technologies like LLMs, machine learning, and computer vision on platforms such as Google Cloud Platform (GCP). Sharad has a proven history of architecting and deploying end-to-end AI-powered solutions.
Sharad is a distinguished technology leader renowned for driving digital transformation and groundbreaking innovation, with a profound specialization in Artificial Intelligence. He is at the forefront of AI research, particularly in the development of advanced AI agent frameworks, leveraging technologies like LLMs, machine learning, and computer vision on platforms such as Google Cloud Platform (GCP). Sharad has a proven history of architecting and deploying end-to-end AI-powered solutions that directly address complex business challenges, significantly boosting project profitability (e.g., through his work on agentic AI bots). His expertise extends to architecting contact center technologies powered by GCP Generative AI, showcasing his ability to bring bleeding-edge AI to practical, high-impact applications. Crucially, his comprehensive background includes leading multiple software testing projects on complex ERP systems (like SAP and Oracle), demonstrating a deep commitment to quality assurance, defect reduction, and efficiency gains (achieving 2-5x improvements). This holistic approach ensures the robust and reliable delivery of cutting-edge AI solutions, making him a highly sought-after advisor and team builder.
All Articles by Sharad Nagariya
All Stories by Sharad Nagariya
|
A Proactive Framework for Testing AI Agent and Application Quality Agentic AI's autonomy creates new security vulnerabilities that traditional methods cannot handle. Its independence expands the attack surface, allowing compromised agents to access sensitive data. Threats like prompt injection, which hijacks an AI's logic with malicious inputs, and chain-of-thought exploits that manipulate its reasoning, require proactive, adaptive defenses. Security must be designed into these systems from the ground up, using tools like agent-level firewalls and continuous behavioral monitoring to secure this new class of software. |