How chain data is shaped, and how to index, stream, store and model it. 14 chapters, 23 hands-on labs, 3 portfolio projects and a private GitHub repo of runnable code.
PySpark from page one. Ten chapters that take a Python user who knows pandas and turn them into someone who can write, read, and debug production PySpark, without a three-chapter detour through distributed-computing theory.
Running ClickHouse at petabyte scale takes more than knowing SQL. This practical guide shows experienced engineers how to design, deploy and operate production clusters, from data modeling and ingestion to Kubernetes, query tuning, observability, disaster recovery and cost control, with real-world patterns and concrete configurations throughout.
Your CPU is not slow. Your data is in the wrong place. Every cache miss your hot loop pays is not a hardware limitation, it is a layout decision that was never made consciously. The pointer your code chases, the struct field your loop skips over, the heap allocation that landed three megabytes from the last one: these are the real performance bottlenecks in CPU-bound systems, and none of them show up in a complexity analysis. This book teaches you to see your data the way the hardware sees it, measure what it costs, and redesign it so the memory system works with your code instead of against it.
PyTorch Deep Dive is a practical guide to mastering modern deep learning with PyTorch. From core concepts to advanced topics like transformers, diffusion models, and production deployment, it combines clear explanations, hands-on examples, and real-world best practices to help you build and scale AI applications with confidence.
SQLite is everywhere but running it well in production takes more than flipping a few settings. This book goes beyond the basics to show how SQLite really works, then walks through the patterns, tools and tradeoffs behind reliable, high-performance deployments. Packed with practical examples, it's built for engineers who want confidence in production.
Data platform interviews test CDC, streaming, warehouse modeling, and trade-offs two levels deep — no generic system-design book covers it. This one does: 50 full mock interviews, 122 rapid-fire Q&A, and the framework that turns a vague prompt into a hire.
Explore data before you trust it. Data Exploration and Visualisation with R shows how to uncover structure, diagnose data-quality problems, challenge misleading patterns, and turn imperfect observations into credible evidence through thoughtful, reproducible visual analysis.
AI can analyze the data. But who decides what the data means—and what to do next?Modern organizations have more information than ever, yet making good decisions remains difficult.AI-Powered Smart Decision Making shows how to combine AI, data, analytics, and human judgment to turn information into insight and insight into action.Discover a practical framework for making better decisions across strategy, marketing, sales, operations, finance, risk, and beyond.AI is not the decision. AI improves the decision process. Table of ContentsIntroduction—The New Era of AI-Powered Decision MakingChapter 1 — The Foundations of Better Business DecisionsChapter 2 — From Intuition to Data-Driven Decision MakingChapter 3 — Understanding Data for Better DecisionsChapter 4 — Business Analytics: Turning Data into InsightChapter 5 — How AI Changes Business Decision MakingChapter 6 — Building an AI-Powered Decision FrameworkChapter 7 — Asking the Right Questions: AI, Data & Decision ProblemsChapter 8 — Predictive Decision MakingChapter 9 — Prescriptive Analytics and Decision OptimizationChapter 10 — AI for Strategic Business DecisionsChapter 11 — AI for Marketing and Customer DecisionsChapter 12 — AI for Sales, Operations and Supply ChainChapter 13 — AI for Financial and Risk DecisionsChapter 14 — Human + AI: The New Decision-Making PartnershipChapter 15 — Bias, Uncertainty and the Limits of AI DecisionsChapter 16 — Trustworthy, Responsible and Explainable AI DecisionsChapter 17 — Measuring Decision Quality and Business ImpactChapter 18 — Building an AI-Powered Decision CultureChapter 19 — Designing an AI-Powered Decision SystemChapter 20 — The Future of Smart Decision MakingConclusion — From Data to Better DecisionsDisclaimer
The Polymorph Data Language (PDL) is a versatile, compact and fast alternative to JSON, YAML, XML and CSV. PDL supports streams of fields, objects, compact objects, cyclic object graphs, tabular data, table trees, and custom instructions. PDL supports this with a syntax that is simple to tokenize and parse and which gives good performance during tokenization and parsing. PDL's custom instructions can be used to implement a whole Scheme-like programming language, should you want to.
Stop only using databases — start understanding them. A living study of PostgreSQL internals for backend engineers: how pages, indexes, and WAL actually work, written clearly while learning, not after mastery.
A Qlik application can show an answer that belongs to the model rather than to the business, and a fast, familiar screen gives no sign of it. The Associative Analyst prepares you for the Qlik Sense Business Analyst and Data Architect certifications and teaches you to prove what an application displays, and what a pass on either examination does and does not prove.
A query can be correct and unaffordable, a metric can mean two things in two tools, and a permission can outlive its purpose. Analytics on Google's Terms prepares you for the Google Cloud Associate Data Practitioner exam and teaches the three questions, who pays, who defines and who can see, that decide whether an organisation can trust the way you use BigQuery, Looker and Data Studio.
Tableau makes it easy to build a clean, persuasive view whose number is wrong, and neither certification asks whether you can defend it. The Tableau Practitioner prepares you for Salesforce Certified Tableau Desktop Foundations and Salesforce Certified Tableau Data Analyst, and teaches you to say what every filter, calculation and encoding claims, and whether the claim holds.
An analyst certified on only the delivery layer or only the engineering layer of Microsoft's analytics stack is structurally blind to the half that the other credential governs. Power and Platform prepares you for PL-300 and DP-600 as one workflow, and teaches you precisely what passing both examinations does and does not prove.