About LakeBench

A place to practise data engineering, not just read about it.

You write SQL, Python, Pandas and PySpark in this browser tab, run it on real tables, and find out straight away whether it is right. There is nothing to install, no cloud account to create and no cluster bill.

  • Runs in your browser
  • Real tables
  • Instant feedback
12
learning tracks
224
lessons
386
coding problems
1,034
interview questions
56
case studies

Counted from the live site when this page was built.

How every exercise works

Lessons, problems, case studies and the 45-day plan all use the same loop. You write a small piece of code, run it on a prepared table, and a checker compares your result with the expected one. When it does not match, you see what differs, change one thing and run again.

That loop is the point. Reading an explanation feels like learning; getting a wrong answer and fixing it is where the learning actually happens.

Write your code, run it on real tables, check it against the expected answer, then try againwrong? read the diff, fix one thing, run againWritequery or scriptRunon real tablesCheckagainst the answer

Who it is for

Students, freshers and people switching careers who need a path from “what is a table” to warehouse SQL, Python scripts and pipelines. It also suits working engineers who want practice on the things that break in real jobs: duplicate rows after a join, late data, slow queries and SCD Type 2 history.

What it is not

Finishing lessons is practice, not a qualification, and it does not guarantee a job. LakeBench does not publish hiring numbers, alumni counts or testimonials it cannot show you. Reviews on the site are written by signed-in learners and approved before they appear.

Who builds it

LakeBench is built and run by one person, K Hunnurji Rao, a senior data engineer with 5+ years in the field. I built it for people learning data engineering or entering the data field, so they can practise and track their interview preparation in one place. It started from a simple frustration: most interview preparation is either algorithm puzzles that rarely come up in a data job, or videos where you watch someone else type.

The lessons, problems, datasets and checks are written for this site. Scenarios are modelled on the kind of work data teams do, with made-up companies and data. Nothing is copied from company question banks or material under an NDA.

Free and Pro

The core lessons in every track, the beginner problems, the first days of the 45-day plan and the daily problem are free. A Pro pass (₹399 for 30 days) unlocks the advanced modules and the full problem library. It is a one-time payment for a fixed number of days and does not renew on its own. See pricing for the passes.

Contact and support

Message me on WhatsApp, email [email protected], or use the contact form. Passes can be refunded within 7 days if unused, as the refund and cancellation policy explains. LakeBench is operated by K Hunnurji Rao.