AI & Agentic Systems

I am currently working on Agentic Commerce initiatives involving MCP (Model Context Protocol), UCP, and AP2, where the objective is to design intelligent agents capable of contextual decision-making, structured tool invocation, and multi-step workflow orchestration within secure commerce ecosystems.

My experience includes building and managing AI agents with tool-calling capabilities, handling agent run lifecycles, implementing prompt engineering strategies for predictable and structured outputs, and integrating LLM services into distributed Azure-based architectures. I work extensively with Microsoft AI Foundry to design persistent agents, manage execution states such as tool invocation flows, and orchestrate AI-driven interactions with backend systems. This includes programmatic handling of agent state transitions, external tool execution, and reliable system integration.

In addition to backend AI systems, I have implemented chatbot solutions and conversational AI interfaces that enhance user interaction while maintaining reliability, security, and auditability. My broader focus lies at the intersection of AI, cloud-native architecture, distributed systems, and regulated domains such as payments, where intelligent systems must operate with high standards of compliance, observability, and production readiness.