I'm an MBBS graduate moving into clinical research and health data roles. I learn by building small projects that follow how real teams work: a written plan first, checks against known answers, and honest limits.
| Project | What it shows |
|---|---|
| drug-adverse-event-explorer | Drug safety: public FDA adverse event data, drug comparison, PRR signal check |
| clinical-data-management | Edit checks, a query log, simplified SDTM-style datasets and a dashboard on a synthetic trial |
| clinical-trials-tracker | Clinical operations: recruiting trials, stale records and a generated report from ClinicalTrials.gov data |
| clinical-research-analysis | Survival analysis, Cox regression and a small meta-analysis, written up as a paper |
| healthcare-sql-analysis | SQL (JOINs, CTEs, window functions) on a synthetic hospital database |
Two projects use synthetic data, and the lung cancer analysis uses a published teaching dataset. No patient information is involved in any of them.
Python (pandas, statsmodels, matplotlib, Streamlit), SQL, Git and GitHub, data validation, survival analysis, meta-analysis and report writing. Concepts I have worked with: pharmacovigilance signal detection (PRR), clinical data management, and simplified CDISC SDTM.
Entry-level roles in drug safety, clinical data management, clinical operations or evidence and biostatistics support, in [city or India].
GCP training and R.