Priya Sharma

Volunteering with a data-for-good group taught me more SQL than any course. If you're in Data Science, look for nonprofits that need simple dashboards. It looks great on a CV too.

My full path from SIT103 to an analyst internship, for anyone in year 2 Data Science who feels behind. In SIT103 I treated every SQL lab like a mini portfolio piece: I saved the queries in a GitHub repo with a short README explaining the question each one answers. Boring, but recruiters opened that repo more than anything else I linked. Over the break I did the free Kaggle Learn courses on pandas and data cleaning. They take a weekend each and they fill exactly the gaps SIT191 leaves. Then I picked ONE finished Kaggle competition, a retail sales forecast, and spent three weeks on it. I did not care about the leaderboard. I cared about writing a clear notebook: what the data looked like, what I cleaned and why, which features mattered, and what I would try next. That notebook became the centre of every interview. Two interviewers literally opened it and asked me to walk through a cell. The SQL questions in the technical round were the same joins and window functions I had practised in SIT103. If you only do one thing this trimester: make your SQL labs public and write one honest Kaggle notebook. Depth beats a long list of certificates.

Canva is opening data analyst intern and graduate roles soon and I am happy to refer Deakin Data Science students. What I look for before I refer: one SQL project with a README that explains the business question, and one dashboard (Tableau or Power BI) someone non-technical can read in 30 seconds. SIT103 and SIT191 work counts if you polish it. Reply with links, and I will give short feedback to everyone, referral or not.

The Canva analyst interview was 70% SQL. Not fancy SQL, just clean joins, window functions, and explaining your reasoning out loud. Mode's free SQL tutorial covers all of it. SIT191 theory came up exactly once.

Kaggle advice for students: don't chase leaderboard positions. Pick one finished competition, write up your approach as a blog post, and put THAT on your resume. A documented 60th percentile beats an unexplained bronze.

neetcode.ioNeetCode 150Curated interview problems with free video explanations.

Honest SIT191 review: the content is dry but it is load-bearing for the whole degree. What saved me was redoing every workshop exercise in Python with pandas afterwards. Twice the work in year 1, half the work every year after, because every later unit assumes you can wrangle a dataframe. StatQuest on YouTube explains the theory better than the lectures, so watch the video BEFORE the lecture and the lecture becomes revision. The exam is formula-sheet friendly; make your own sheet, don't download one.

Tableau Public is free and recruiters actually click the links. Two polished dashboards on real datasets (not iris, not titanic) did more for my applications than my entire academic transcript. Here's the one that got the most questions in interviews.