I’m an MSc Economics graduate from GIPE, Pune. Before that, I studied at St. Anthony’s College in Shillong. This portfolio is a record of three projects I worked on over the past year — each one taught me something different about what actually makes analysis useful.
Tools: SQL
Started with a straightforward question: what separates borrowers who repay from those who don’t? I pulled borrower-level data and looked at income, loan size, and repayment history together rather than in isolation.
The first segmentation I built was technically clean but hard to explain. I simplified it — fewer buckets, clearer logic — and that version ended up being more useful. Interpretability mattered more than sophistication here.
Tools: SQL, Python, Excel
RFM (recency, frequency, monetary value) is a well-known framework, but applying it to real transaction data is messier than it looks. Standard thresholds gave me segments that looked fine on paper but didn’t reflect actual customer behaviour.
I iterated on the cutoffs, cross-checked segments against real purchase patterns, and ended up with a structure that made intuitive sense. The adjustment process was more instructive than the final output.
Tools: Excel
This one started as an overcomplicated mess. I had too many variables and too many assumptions chained together — the model was fragile and hard to interrogate.
I stripped it back to a supply-demand balance sheet: production, inventory draws, and demand assumptions linked to a directional price bias. It doesn’t give you a precise number, but it gives you a clear view of which way pressure is building. That’s usually what you actually need.
My academic focus has been on econometrics, public finance, and financial statement analysis. I interned as a Data Analytics Trainee at MedTourEasy, where I worked with Excel and SQL on healthcare data. Primary tools: SQL, Excel, Python — I’m comfortable with Python for supporting analysis, though I don’t use it as a primary tool yet.
I was selected for the McKinsey.Org Global Leadership Program (one of 200 globally in 2024) and the OMLAS Fellowship — both of which pushed me to think about analysis in terms of decisions and outcomes, not just outputs.
Mohit Das mohitxdas@gmail.com LinkedIn · GitHub · Live Portfolio