Publications
Writing & Research
Engineering-first articles and research papers on production-grade AI architecture, agent security, and performance engineering.
Industry
Technical Articles
Practical, engineering-first articles published on Microsoft Tech Community. Designed to help teams ship reliable systems under real constraints.
The Hidden Boundaries of Modern AI
A systems-level look at the practical constraints shaping modern AI - from context, memory, and execution limits to the engineering boundaries teams must design around.
Microsoft Tech CommunityBuilding an Auditable Security Layer for Agentic AI
A practical architecture for adding auditable controls, governance, and security boundaries around agentic AI systems before they reach production.
Microsoft Tech CommunityThe Hidden Architecture of Nano-Architectures
A deep dive into why your LLM's output is a property of the executed plan, not the checkpoint — how continuous batching, VRAM headroom, cache residency, kernel tactics, and reduction order shift under load and can flip early tokens into full semantic drift.
Microsoft Tech CommunityThe Hidden Memory Architecture of LLMs
Exploring the internal memory dynamics and kernel-level behaviors that drive large language model inference.
Microsoft Tech CommunityAI Didn't Break Your Production — Your Architecture Did
Why production AI failures are architecture problems, not model problems. A deep dive into resilient system design.
Microsoft Tech CommunityZero-Trust Agent Architecture: How To Actually Secure Your Agents
A comprehensive framework for building secure, governed agent systems with zero-trust principles.
Microsoft Tech Community.NET 10 and Memory: Less Heap, Smarter GC, Faster Apps
Performance engineering deep dive into .NET 10's memory improvements and their impact on real-world applications.
Research
AI Research Papers
Academic and applied research in medical imaging AI, NLP, and computer vision. Published in leading journals including NIH.
AI Research & Security: LAAF - Logic-layer Automated Attack Framework
An academic paper introducing LAAF, a logic-layer automated attack framework for evaluating and strengthening AI security controls.
MelaVision
A machine learning model for early melanoma detection using computer vision and deep learning.
Covid-19 Classification
A deep learning model for classifying Covid-19 cases using medical imaging, published on NIH.
DeepInfant
Exploring deep learning applications for infant behavior analysis and monitoring.
Optimal Training Calculator
A tool that helps ML practitioners calculate optimal training parameters for deep learning models.
Stock Risk Analyzer
A machine learning-powered tool for analyzing and predicting stock market risk levels.
Collaborations