PinnedArmin Norouzi, Ph.D·Aug 6Introducing Data-Driven Personal FinanceHey friend,A response icon1A response icon1
PinnedInData Science CollectivebyArmin Norouzi, Ph.D·Jun 5Backpropagation Without the Magic: A First-Principles DerivationEvery ML engineer uses backpropagation daily. Most treat it as a framework primitive — loss.backward() runs and gradients appear. That…
PinnedInArtificial CornerbyArmin Norouzi, Ph.D·Sep 12, 2023Retrieval-Augmented Generation (RAG): A Short IntroductionDive into the essence of RAG in AI: a tool amplifying data richness & optimizing outcomes by connecting AI models to vast data resources.
PinnedInLevel Up CodingbyArmin Norouzi, Ph.D·Jan 22, 2024Diagram As Code: Crafting AWS Architecture Diagrams Using PythonDiscover the art of diagramming AWS infrastructures using Python’s ‘diagrams’ library.A response icon7A response icon7
PinnedInLevel Up CodingbyArmin Norouzi, Ph.D·Dec 12, 2023Unit Testing at Google: Building Robust Software with Strategic TestingExplore the nuances of unit testing in Google’s software engineering landscape.A response icon4A response icon4
InData Science CollectivebyArmin Norouzi, Ph.D·6d agoAI Reasoning vs. Human Thinking: What’s the Difference?Does AI truly understand? See how LLMs reason on GPUs (KV cache, FlashAttention, GRPO) and how that compares to the neuroscience of the…A response icon1A response icon1
InDataDrivenInvestorbyArmin Norouzi, Ph.D·Oct 19The Engineering Proof of the 4% Rule: Lyapunov Stability for Retirement WithdrawalsThe 4% rule is the most important constant in retirement planning, and almost nobody checks it the way an engineer checks a constant.
InTowards AIbyArmin Norouzi, Ph.D·Aug 24The Evaluation Stack: Metrics That Predict Production QualityThe hardest engineering problem in LLM deployment is not latency or cost — it is knowing whether your model got better. Metrics that seem…
InTowards AIbyArmin Norouzi, Ph.D·Aug 18Tool Call Orchestration: Sequential, Parallel, and DAG ExecutionA research agent calling 6 tools sequentially waits 1,832ms at P50. The same 6 tools run in parallel — ignoring all dependencies — finish…
InLevel Up CodingbyArmin Norouzi, Ph.D·Aug 17Open Source vs. Closed API LLMs: A 2026 Production Decision GuideThe open vs. closed debate in LLMs has been running since GPT-3 launched in 2020, but in 2026 the conversation has fundamentally changed…