Sub-Millisecond Feature Stores: Architecture & Latency Bounds
How to design unified streaming feature ingestion with Kafka and Redis that satisfies sub-5ms p99 SLAs while eliminating training-serving skew.
Read the articleTechnical Journal
In-depth technical writing on sub-millisecond feature stores, probability calibration, Triton/Ray inference serving, and in-process analytics with DuckDB.
How to design unified streaming feature ingestion with Kafka and Redis that satisfies sub-5ms p99 SLAs while eliminating training-serving skew.
Read the articleAll notes
Sample articles for this portfolio template.
Why modern deep neural networks produce overconfident probabilities, and how temperature scaling and isotonic regression prevent catastrophic errors.
Read article
Architecture Guide
A hands-on guide to containerizing complex ML ensembles with NVIDIA Triton, dynamic batching, and Ray Serve concurrency controls.
Read article
Engineering Analysis
For datasets under 500GB, spinning up multi-node Spark clusters introduces massive operational friction. How DuckDB enables ultra-fast in-process analytics.
Read articleTry another topic or reset the filter to read all articles.
Demo content. These articles are sample writing for the portfolio template, not published editorial work. Replace them with your own notes before publishing.