data analyst · network analytics
Turning messy KPI dumps into
decisions people trust.
Dashboards, clustering pipelines and statistical studies for a national telecom network. Charts below are drawn from the real analysis output; operator and competitor names are redacted to Operator A/B/C.
Selected work
About
I work where data engineering meets network engineering: Ookla/OpenSignal-style KPI scoring, RAN health metrics, change-point detection, clustering, forecasting. Python + SQL + DuckDB daily; Streamlit when someone needs to click on it.
- python
- sql
- duckdb
- pandas
- sklearn
- streamlit
- statsmodels
- prophet
- docker
Numbers and charts are real output from these projects. Operator names, site IDs
and credentials are not published; the redaction happens at build time in
build_data.py and build_shots.py.