Somesh Zanwar.

Experience

Where the habits came from

Work

Junior Data Analyst

IVS Software Solutions · Pune, India · Sep 2022 – Sep 2024

Figures below are from my resume and self-reported. Employer data is not public.

  • Analyzed 80,000–150,000 operational records per reporting cycle with SQL and Python. Validation, reconciliation and discrepancy resolution improved reporting accuracy by 30%.
  • Automated recurring reporting, saving 4–6 hours per cycle, and optimized SQL joins and incremental loads, cutting data-preparation time by 40%.
  • Worked with three internal teams on Power BI dashboards, KPI definitions, reusable datasets and recurring reports.
  • Documented metric definitions, transformation logic, validation procedures and discrepancy resolutions so recurring reports stayed consistent.
Education

M.S. Data Science

The University of Texas at Arlington · Jan 2025 – Dec 2026

GPA 3.67 / 4.00. Capstone with UTA's Office of Information Technology (in progress, private).

B.E. Artificial Intelligence and Machine Learning

Savitribai Phule Pune University · Aug 2020 – May 2024

GPA 8.2 / 10.

ToolsListed with where I used them
Querying and modellingSQL, PostgreSQL, dbt, DuckDBEvery project above. dbt models and tests are the evidence source in MetricTrust and DataTrust OS. SQL join and incremental-load tuning cut preparation time by 40% at IVS.
Analysis and experimentationPython, pandas, NumPy, SciPy, A/B testing, hypothesis testingDecision Intelligence Experimentation Platform (z-test, lift, confidence intervals, segment analysis) and Metric Decomposition Engine.
BI and reportingPower BI, Tableau, KPI definition, stakeholder reportingPower BI governance dashboard on the AI Data Governance Platform. Dashboards and KPI definitions with three teams at IVS.
Services and testingFastAPI, SQLAlchemy, pytest, Docker, Git, CIMetricTrust Control Plane (21 tests, signed permits) and DataTrust OS (Python tests in GitHub Actions).
Machine learningscikit-learn, XGBoost, SHAP / LIME, cross-validationCoursework and the in-progress capstone. Not the centre of my public projects, so I have not dressed it up as one.
Governance and trustData quality checks, lineage, policy controls, audit evidenceDataTrust OS, Data Quality-Aware Agent Governance and the AGT contributions.