1d ago · 9 min read · Building the first version of an AI workflow is usually easy. Connect an LLM to a few tools. Add some instructions. Let the model decide what to do next. Run the demo. It works. The problem starts l
Join discussion
3d ago · 11 min read · 🗓️ Last updated: May 2026 — reflecting current production patterns and the OWASP LLM Top 10 (2025). Every few years, the software industry rediscovers the same lesson. A new architectural pattern ar
UAlok and 1 more commented
May 23 · 13 min read · Recently, I’ve been exploring a lot about RAG systems and local LLMs, not from a research perspective, but more from a curiosity perspective as a developer who genuinely enjoys building things and und
Join discussion
May 21 · 6 min read · TL;DR PyTorch’s DataLoader can be 50-124x slower than direct tensor indexing for in-memory GPU workloads. We reproduced a real PyTorch issue on an RTX 4090 and traced every CUDA API call and Linux ke
Join discussion
