Blog

Tech outlets love flashy benchmarks, but most miss the fundamental principles of testing with controls. Here’s why their comparisons are often meaningless.

While UV promises impressive speed improvements and there is a host of tech influencers now migrating to UV, Poetry still remains the more reliable, mature choice for Python package management in 2025.

Why I moved from LMStudio to Ollama and setting up OpenWebUI as frontend.

Looking at the current state of agents, It is possible to replicate them with a simple loop.

Adapting Shape Up methodology for data science work, what works and what doesn’t

Key environment settings for running Ollama efficiently on Windows and Linux.

Selecting and enabling which GPUs are visible to ollama within windows.

How the Mac Mini M4 has enabled affordable Local LLM inference and making VRAM-starved NVIDIA cards obsolete for low power and low cost

Around October 2024, The infamous “How many R’s are in strawberry?” question has become a fascinating litmus test for Large Language Models, exposing a fundamental limitation in how these systems process text.

Setting up FUTO keyboard on Android with local Whisper models for private voice typing