Think in Patterns

• 340problems: 300 classic and 40 AI-era
• 36patterns, one chapter each
• 13Foundations topics for beginners
• 913answers, each linked from its exercise and back
• 353Python solution files, run against their tests before release

The problems are free online. Why a book?

• A list gives you questions. This gives you the idea behind them. 340 problems sorted into 36 patterns, each with the words that give it away, so the next problem you have never seen still looks familiar.
• One patient voice, start to finish. Every problem is taught the same way: an everyday scene, the obvious answer, the better one, a drawing, tested Python, why it works, a dry run and the mistakes people make.
• You do part of the work. Questions to predict, dry runs to finish and twists to try, with every one of the 913 answers one link away.
• A route, so you always know what is next. Short on time? Route B is 57 problems that cover every classic pattern at least once: at three a day, under three weeks.
• Checked before it reaches you. Every one of the 353 Python solution files runs against its tests before a release.

And you can check all of it before paying: read 42 pages free, try the free lessons, and if you buy and it is not for you, there is a 7-day refund.

One problem, start to finish

1. An everyday sceneThe problem told as something you already know.
2. The obvious answerBrute force first, and what it costs.
3. The better oneThe idea that removes the wasted work.
4. A drawingThe key moment, drawn by hand.
5. PseudocodeThe plan in plain words.
6. Tested PythonReadable code that runs against its tests.
7. Why it worksThe reason, not only the steps.
8. A dry runThe code traced on a small input.
9. The mistakes people makeThe traps, before you fall in.

And you do part of the work

Reading a solution feels like understanding it. These three make sure it is.

• Predict first

  One question to answer before you read on. The next sections tell you if you were right.

• Your turn

  After the dry run, run the method by hand on a new input. The → answer link takes you there and ↩ brings you back.

• A twist to try

  The same problem with one rule changed. Adapt the solution yourself, then check its answer.

Every answer is printed at the end of its chapter, never next to its question, so you can’t peek by accident.

You don’t read 2,288 pages. You follow a route.

Each route has its own page in the book. Every step on it is a link, so you always know what to read next and when you are ready to move on.

• Route A · Foundations first

  Take it if data structures are new to you, or “hash map” and “O(n log n)” still feel vague.

  It starts with the 13 Foundations topics: what each structure is, what it costs and where real software uses it. About two weeks at one topic a day.

• Route B · Interview core

  Take it if you know the basic structures and have an interview coming.

  57 problems that unlock the most others, covering every classic pattern at least once. At three a day, under three weeks.

• Route C · Applied (AI-era)

  Take it if you are aiming at ML, data, backend or AI-agent roles.

  The core problems the applied chapters build on, then the six applied chapters: NumPy, pandas, embeddings, APIs, async code and AI agents.

Not sure? Try the first problem without reading its solution. If you can’t write working code in 20 minutes, take route A. If the answer comes to you quickly, take route B.

What you need before page 1: basic Python (variables, if, for and while loops, functions, lists and dicts, running a .py file). No maths beyond school algebra, and no computer-science degree.

Look inside

Nine pages from the book, exactly as printed. Turn them as you would the real thing, and open one larger to read it.

p. 13 · Three ways through the book, and what you need before page 1p. 17 · Route B: the 57-problem Interview core, every step with its page

What’s inside

The 36 patterns

300 classic interview problems in 30 patterns, then 40 AI-era problems in 6 more. The number is how many problems each chapter has.

Around the problems

Everything else that turns 340 problems into a course.

• 13 Foundations topics

  For beginners: What DSA is, and why interviews ask it · Big-O without fear · Memory, variables and references · Arrays and Python lists · Strings · Hash maps · Linked lists · Stacks and queues · Recursion · Sorting and searching · Trees · Heaps and priority queues · Graphs.

• The interview toolkit

  The six steps to follow out loud in every interview, a cheat sheet from the words in a problem to its pattern, Big-O and Python at a glance, and what to expect in online assessments.

• Interviews in the AI era

  How AI-assisted rounds work, and what to check in an assistant’s code. Then 40 applied problems: NumPy, pandas, embeddings, APIs, async code and AI agents.

• Practice that sticks

  A checkpoint after every chapter, 4 mixed practice sets, 4 timed coding rounds and 6 projects.

• Plans and a tracker

  8-week and 14-day plans, and a tracker with re-solve dates at 3, 10 and 30 days.

  Solved it once? Solve it again after 3, 10 and 30 days.

• A fully linked PDF

  2,288 A4 pages with contents, index, every “see page” reference and bookmarks for every chapter and problem. Plus 36 chapter PDFs, each with its own answers, handy on a phone.

Try 42 pages free

Who it’s for

• For you if

  You know basic Python (loops, functions, lists, dicts) and want to pass coding interviews: students, career changers and working developers. No computer-science degree needed.

• What it is not

  Not a system-design or behavioural-interview book. Not a promise of a job or an interview result. Not affiliated with LeetCode or any company named in it: LeetCode numbers are there only so you can find each problem to practise on.

Get the notebook

Questions

About the author