Quick context before I get into it :— second year at IIT Guwahati, MnC branch, applying for SWE internships. I had one SQL course and one lab before this. Everything else I'd be figuring out while building.
-I want to be upfront: I haven't built this yet. I'm posting before I start because I've seen people waste time on projects that they later think are useless. So help me if possible.
What I'm planning ::-
-An analytics engine on the complete IPL ball-by-ball dataset (~200k delivery records, 2008–2024). The angle is SQL depth — not a dashboard, not a MERN app, just genuinely hard database work.
The things I want to implement:
Window functions — rolling batting form with ROWS BETWEEN frame clauses, milestone tracking, partnership analysis using LAG()
Recursive CTE for winning streaks (gaps and islands problem)
AFTER INSERT trigger keeping a pre-aggregated stats table in sync so career totals don't need full recomputation on reads
EXPLAIN-based index optimization with actual before/after benchmarks
A small Python pipeline pulling from Cricsheet's raw JSON (not the cleaned Kaggle CSV) with extract, validate, load steps
Four FastAPI endpoints wrapping the stored procedures so there's something you can actually hit
Why I think this might be worth it
-- The stuff I listed above — recursive CTEs, trigger-based materialization, covering index strategies — I don't see that often in undergrad portfolios. I might be wrong about that.
Why I'm not sure
The IPL dataset is genuinely overused. I know that. The Cricsheet pivot helps a bit since the raw JSON needs actual transformation work, but it's still cricket data which every analytics tutorial seems to use.
I'm also not sure whether "deep SQL" reads as impressive to an SWE hiring manager or whether they just want to see another React project.
:::so help me to decide weather this is worth keeping in resume:::
What I'm actually asking
Is SQL depth a differentiator in 2025 intern resumes or is it invisible to most SWE hiring managers? Would you shortlist this or skip it?
If the consensus is skip it, I'd genuinely rather know now.
Top comments (0)