proomt

Search

Search posts, papers, and topics

Hall of Fame

Hall of FameMartin Zinkevich201643 min readtutorialintermediate

Rules of Machine Learning

Read on google.comHN 84·7 comments

Summary

Google’s “Rules of Machine Learning” is a practical style guide for engineers building ML‑enabled products. It stresses solid pipelines, early metric instrumentation, simple first models, and reusing heuristics as features.

  • Don’t launch ML until you have data; simple heuristics can deliver ~50% of the benefit.
  • Instrument metrics and set up experiment frameworks before building the model.
  • Start with a simple model and a reliable pipeline; test data flow and serving separately.
  • Turn existing heuristics into features or preprocessing steps rather than discarding them.

ML engineers and product teams should read it to avoid common engineering pitfalls and accelerate reliable model deployment.

6/10
machine learningmlpipelinesmetricsfeatures

Related reading

  1. Hidden Technical Debt in Machine Learning Systems

    This paper introduces the concept of technical debt in machine learning systems, arguing that ML systems accrue unique and significant maintenance costs beyond traditional software engineering. It identifies several ML-specific risk factors like entanglement, hidden feedback loops, and data dependencies that erode system boundaries and increase long-term operational expenses.

    Hall of Famenips.cc25 minpaper

  2. Changing the game: How Google uses agentic AI to secure hundreds of millions of lines of code

    Google’s AI & Infrastructure team built an agentic pipeline (Mantis) that runs pre‑submit AI‑driven scans on every code check‑in, validates findings with a fast triage agent (AST + call‑graph analysis) achieving >92% precision in <1 min, then auto‑generates fixes via a bug‑fix agent. Localized threat models and a two‑step scan cut false‑positives to ~3% and prevent hundreds of vulnerabilities eac…

    Google Cloud Bloggoogle.com4 min

  3. Cloud CISO Perspectives: How Google monitors AI threats and advances AI defenses

    Google’s Threat Intelligence team outlines three AI‑driven shifts—software build changes, expanded attack surface, and enhanced threat capabilities—then describes their multi‑model, graph‑based defense stack (AI Threat Tracker, in‑editor “spellcheck”, Wiz Security Graph, Gemini‑powered AI Threat Defense) and concrete threat examples like supply‑chain poisoning, LLMJacking, and AI‑orchestrated cre…

    Google Cloud Bloggoogle.com11 min

  4. AI Norms & Values, Part 1 of 3: How We Do Business at Honeycomb

    Honeycomb’s first AI‑norms document (How We Do Business) codifies eight concrete GTM principles – value‑first, human‑centric, truthful, earned‑asks, hard‑but‑fair competition, respect for attention, AI‑augmented responsibility, and durable relationships – and argues that AI merely amplifies existing culture rather than creating new rules.

    Honeycombhoneycomb.io7 min

  6. Scaling Telco Autonomy: Leveraging GNNs with Distributed GraphFlow

    Google Cloud’s blog introduces Distributed GraphFlow (DGF), an open‑source Python library for building and scaling Graph Neural Networks (GNNs) on a Spanner‑backed digital twin of telecom networks. The post outlines the three‑layer architecture (digital twin on Spanner Graph, ML layer with DGF, AI agents) and highlights DGF’s high‑level API (5‑line example) and low‑level primitives, but provides…

    Google Cloud Bloggoogle.com3 min