Blogs and Wiki

Systematically determine the optimized spot where we can push our hardware to the fullest possible use. Multithreaded optimization depends on multiple factors including CPU/GPU type (M4Max vs 14900 or MetalCUDA. Operations involving space-time tradeoffs like heap usage need to be fine tuned around batch sizes.

Drone API for Entity Extraction

Machine Learning - AI - Tensorflow - Keras - NVidia - Google

Various Kubernetes cluster installations

Get a private deployment of Cohere Command models

gcp-infrastructure-as-code

Java Spring boot canary for testing REST workflows

Conslidated private data center artifacts

Tensorflow 2 Java API for CUDA and Metal