I am an AI engineer and product owner with over seven years of experience across enterprise software, data engineering, and applied AI. At Cesta, Inc., supporting an Accenture engagement, I work on generative AI and agentic AI systems, combining hands-on engineering with technical leadership, product ownership, and cross-functional delivery.

My work spans large language models (LLMs), multi-agent systems, retrieval-augmented generation (RAG), vector databases, and tool-using agents. I build workflows using Temporal, integrate enterprise tools through the Model Context Protocol (MCP), and connect models through LiteLLM. My experience includes persistent conversational state, real-time streaming with server-sent events (SSE), and coordinating agents across planning, code generation, building, and testing.

I focus on the engineering foundations that make AI applications dependable: evaluation, observability, security, and workflow recovery. My work and technical interests include LLM-as-a-judge, RAG evaluation, OpenTelemetry instrumentation, distributed tracing, and monitoring through logs, metrics, and traces. I explore how tools such as HyperDX, ClickHouse, Prometheus, and Grafana can help teams understand agent behavior, investigate failures, and measure latency and reliability.

Security and AI governance are a integral part of my work. I implement guardrails to mitigate prompt injection, sensitive-data exposure, and unauthorized tool use. My approach includes Microsoft Presidio for identifying and redacting personally identifiable information (PII), Open Policy Agent (OPA) for policy-based authorization, and Keycloak, JWT validation, and role-based access controls for securing multi-tenant applications. I focus on combining input and output validation, access policies, and controlled tool execution to strengthen AI systems against adversarial behavior.

My technical background spans Python, FastAPI, SQL, PySpark, Snowflake, BigQuery, PostgreSQL, MongoDB, and Milvus, complemented by knowledge of Snowflake, Databricks and generative AI fundamentals. I bring experience with AWS and Google Cloud, along with a foundation in Oracle Cloud Infrastructure, ETL pipelines, cloud data migration, API development, and data quality at scale. My approach to engineering delivery draws on Scrum, product ownership, and Lean Six Sigma principles to connect technical execution with business priorities and continuous improvement.I have implemented multi-tenant authentication, Keycloak-based identity integration, JWT validation, and role-based access controls. I also explore agent frameworks and deployment approaches, with an interest in interoperability, AI governance, and production readiness.

Alongside engineering, I translate business goals into technical requirements, prioritize product capabilities, coordinate delivery across teams, and connect architectural decisions with customer needs. I hold a master’s degree from North Carolina State University and serve as a reviewer for the NeurIPS 2026 JUDGe Workshop.

I make complex AI architectures accessible through practical examples and engineering lessons. My sessions connect the details of building and evaluating AI agents with the product and leadership decisions needed to bring them into real-world use.