MediCheck
MediCheck is a Python, Tkinter/ttk, and MySQL educational clinic-workflow prototype that I originally built in 2024 and substantially redesigned in 2026. The current application follows one continuous workflow from patient check-in and queue management through consultation, transactional visit completion, and historical record retrieval.
Technology: Python, Tkinter/ttk, MySQL

Workflow redesign
I redesigned the original popup-heavy, debug-style interface as a persistent multi-workspace application shell with clear navigation and integrated patient workflows. Consultation is one four-stage experience—Symptoms, Diagnosis, Prescription, and Review—that keeps patient context and confirmed selections visible. This preserved useful backend behavior while improving the existing application instead of unnecessarily rewriting it.
Relational data and SQL
Patients, visits, symptoms, diagnoses, prescriptions, and medicines are represented as related data. Historical records support SQL-driven symptom, diagnosis, and medicine suggestions through relational associations, aggregation, frequency analysis, and ranking. This is not a trained machine-learning model or medically validated AI.
Reliability and state
Visit completion persists the complete logical case in one database transaction. If persistence fails, the transaction rolls back and consultation state remains available for retry. State clears only after successful completion, and duplicate completion is guarded against. The redesign also strengthened validation and state-lifecycle handling.


