HomeData ScienceAI Studio

skills

The working toolkit, honestly framed. Everything listed here is something I have genuinely used, and the last section is what I'm learning right now, listed openly because trajectory matters.

Python & data manipulation

Comfortable with: my daily working environment for analysis and modeling.

  • Python
  • pandas
  • NumPy
  • Matplotlib
  • Plotly
  • Jupyter

Demonstrated in Customer Churn Prediction

SQL & databases

Working knowledge: window functions, CTEs, cohort analysis, and data-quality auditing.

  • SQL
  • SQLite
  • window functions
  • CTEs
  • cohort analysis

Demonstrated in Revenue & Retention Analytics

Machine learning

Working knowledge: supervised learning with a baseline-first, evaluation-focused workflow.

  • scikit-learn
  • logistic regression & gradient boosting
  • cross-validation & temporal validation
  • leakage prevention
  • precision@k, ROC-AUC, log-loss, Brier
  • calibration & error analysis

Demonstrated in Customer Churn Prediction

Statistics & experimentation

Working knowledge: hypothesis testing and knowing when a result is noise. Backed by a mathematics degree.

  • hypothesis testing (chi-square, t-test, z-test)
  • A/B test design & evaluation
  • effect sizes (Cramér's V, Cohen's d)
  • regression

Demonstrated in Revenue & Retention Analytics

Time series & forecasting

Working knowledge: walk-forward backtesting and leakage-safe feature design.

  • walk-forward backtesting
  • lag & rolling features
  • forecast metrics (MAE, WAPE)

Demonstrated in Demand Forecasting

Data visualization & communication

Comfortable with: charts that state a takeaway, dashboards built around decisions, and write-ups a non-technical reader can follow.

  • Matplotlib
  • Plotly
  • Streamlit
  • KPI design
  • written analysis

Demonstrated in Retail Operations Dashboard

Workflow & tooling

Comfortable with: reproducible, version-controlled, tested project work.

  • Git & GitHub
  • pytest
  • virtual environments
  • reproducible pipelines

Currently learning

In progress: listed openly because trajectory matters as much as the current toolkit.

  • PostgreSQL & BigQuery
  • deep-learning fundamentals
  • dbt & data pipelines

Why no skill bars?

Percentage bars and star ratings look precise but measure nothing. Prose framing like “comfortable with”, “working knowledge”, and “currently learning” is what I’d actually say in an interview, so it’s what the page says too.