i build fast.
Applied AI Engineer with 2+ years of experience building production LLM applications, agentic workflows, and RAG systems using Python, FastAPI, LangChain, LangGraph, and Azure OpenAI. Delivered enterprise automation handling 1,000+ monthly requests at 98%+ compliance accuracy, with measurable gains in token efficiency, response time, and release speed.
Engineered a LangGraph, Azure OpenAI, and RAG workflow for 1,000+ monthly ServiceNow requests covering Azure network security changes, reaching 98%+ compliance accuracy. Built retrieval-grounded policy validation with knowledge-graph context, document retrieval, Pydantic structured outputs, and human-in-the-loop review; Redis caching cut token usage by 35% and response time by 45%.
Developed and shipped LangGraph proof-of-concepts featuring token streaming, MCP-based tool integration, and RAG pipelines for internal AI workflow use cases. Built FastAPI-backed prototypes using retrieval orchestration and structured outputs, helping convert exploratory LLM ideas into deployable delivery candidates.
Built and shipped a React Native MVP end to end in under 6 months for a cross-platform startup product, and developed Flask REST APIs supporting 3+ core workflows to improve iteration speed for the product team.
Built an agent-based workflow for 1,000+ monthly Azure network policy operations with ServiceNow integration, retrieval-based context, and policy validation. Added structured execution, auditability, and human-in-the-loop control for production-facing cloud operations.
Built a full-stack dashboard collecting and visualizing data from 500+ AWS resources for cost review and prioritization. Added an LLM-based summarization layer that generated actionable cost-optimization recommendations for underutilized infrastructure.
A conversational AI assistant built with Groq and Oss gpt-20b. KevBot can answer questions, generate text, and tell you about, well, me.
A multimodal AI system designed to unify information retrieval, command execution, and patch management into one interface. Features video captioning, semantic search, PDF/audio RAG, voice-driven execution, and patch management demos.
A modular conversational AI agent built with LangGraph and MCP (Model Context Protocol). PlugGraph allows you to "plug and play" new MCP servers for weather, jokes, info, search, and more.
CRC Press / Taylor & Francis, 2025. Co-authored a review covering speech recognition, neural machine translation, voice cloning/TTS, and lip-sync generation for multilingual video translation.
An article on practical attack surfaces and blind spots in modern AI systems, with security-first perspective.
A piece exploring emotion-aware AI signals and their implications for financial prediction systems.
Listed in the supplied resume as part of the Azure credential set.
Professional credential tied to applied AI delivery and systems work.
Infrastructure and cloud administration credential from the supplied resume.
Coursework: Cloud Computing, Distributed Systems, Data Structures & Algorithms.
I keep learning by trying things, writing about them, and reusing the lessons in the next project.
Let's build better things, together.