Knowledge Space
Workbench
A curated collection of published articles, research explorations, and technical insights — the thinking behind the engineering.
Published Writing
Medium Articles
In-depth technical articles covering LLM optimization, GPU programming, and edge deployment — written for engineers who build.
Semantic Caching with Redis: How to Optimize LLM Cost and Latency
Learn how to reduce redundant LLM API calls using semantic similarity-based caching with Redis, sentence embeddings, and vector search — with a full Python implementation.
Building OpenCV with CUDA Support on Jetson Devices
A practical guide for compiling OpenCV with CUDA support on NVIDIA Jetson platforms, covering prerequisites, CMake configuration, build troubleshooting and verification steps.
Deep Dives
Research & Exploration
Extended research work and technical explorations — from edge inference frameworks to foundational papers.
Adaptive Edge AI Controller
An intelligent edge AI layer for NVIDIA Jetson devices that monitors temperature during real-time YOLO inference and dynamically adjusts workload to reduce throttling, FPS drops, and shutdown risk.
Read Research Article
Adaptive Edge Research Paper
Senior thesis research document
Model Training Research Paper
A benchmarking study where I trained, compared, and evaluated multiple deep learning models.
Social Insights
LinkedIn Posts
Selected technical posts and architecture breakdowns shared with the engineering community.
RAG Architecture Design
End-to-end design of a production-grade Retrieval-Augmented Generation pipeline — covering chunking strategies, embedding model selection, vector store indexing, hybrid retrieval with re-ranking, and prompt engineering patterns for grounded LLM responses.
Next Step
Interested in the projects behind the writing?
The Workbench captures the ideas and research. The deeper implementation work lives in projects — take a look or reach out to discuss systems, AI, and engineering.