Articles
Worth reading
A short take on pieces worth your time — architecture, AI, system design, and engineering careers — each one linked back to its original source.
22 articles
What Is RAG? How Retrieval-Augmented Generation Works
A clear, current walkthrough of retrieval-augmented generation: why it exists (grounding an LLM in your own data instead of relying on what it memorized during training), how the retrieval-then-generate pipeline actually works, and where teams run into trouble in practice — the retrieval step, not the model, is almost always the real failure point. Useful if you're deciding whether RAG is the right pattern for a project, or just want a precise mental model instead of the marketing version.
Netflix Tackles Data Deletion at Scale with Centralized Platform Architecture
InfoQ's coverage of a QCon talk on a problem almost nobody designs for up front: how do you actually delete data, correctly and completely, across dozens of heterogeneous storage systems, at a scale where 76.8 billion row deletions across 1,300 datasets is a normal workload? Netflix's centralized deletion platform is a reminder that "delete" is a distributed systems problem with its own consistency and observability requirements, not an afterthought bolted onto each service. Especially relevant if you're anywhere near GDPR or CCPA compliance work and have been treating deletion as someone else's problem.
Building Effective AI Agents
Anthropic's own engineers stepping back from the hype to draw a precise line between a "workflow" (LLMs and tools wired through code you control) and an "agent" (the model deciding its own next steps) — a distinction that matters the moment you're deciding how much autonomy to actually hand an LLM in production. The five patterns it lays out — prompt chaining, routing, parallelization, orchestrator-worker, evaluator-optimizer — are the ones I keep coming back to when scoping an AI feature, because they're composable primitives rather than a framework you have to buy into. Worth reading before you reach for a heavyweight agent framework: most production use cases turn out to need a workflow, not an agent.
Python 3.13: Free Threading and a JIT Compiler
Real Python's clear-eyed breakdown of the two most consequential changes to CPython in years: an experimental build that makes the GIL optional (real thread-based parallelism for CPU-bound Python, finally) and a JIT compiler that's still early but points at where the interpreter is headed. Neither is production-default yet, and the article is honest about that — it's a guide to what's coming and how to try it, not a victory lap. If you've been telling people "Python doesn't do real threads," read this before you say it again.
Strangler Fig
The pattern for replacing a legacy system without ever staging the terrifying "big bang" cutover: put a facade in front of the old system, and move functionality behind it into the new system piece by piece until there's nothing left to strangle. I've seen a version of this used on more than one banking migration, and the value isn't the metaphor — it's the discipline it forces, routing real traffic through the seam from day one instead of building the replacement in isolation and hoping the switchover goes smoothly. Short, practical, and still the reference point everyone eventually points to.