A role-based care platform where doctors and patients share assessments, therapy activity, events, progress reports and messages, with a rule-based NLP support chatbot.
Project media coming soon
Screenshots and demo videos for this case study will appear here.
The system gives doctors and patients separate experiences. Patients choose a doctor, complete predefined assessments and daily activities, plan events, track progress and message their doctor. Doctors review their patient list, new patients, reports and therapy sessions. A support chatbot answers common questions.
Mental-health care involves ongoing context rather than one-off appointments. The project had to connect scheduling, assessments, activity tracking, reports and communication around each patient.
I built the interface in HTML, CSS and JavaScript, with the application layer in PHP and MySQL across more than 25 endpoint scripts. They cover authentication, email verification and password recovery through PHPMailer, patient assignment, events, activities, progress, reports, therapy tools and doctor–patient chat. The chatbot normalises input with tokenisation, Porter stemming and stop-word removal before matching it against known intents.
This was my first end-to-end system. It gave me practical experience with role modelling, relational data and multi-step workflows, and showed me what I would build differently today.
The chatbot lowercases input, strips punctuation, tokenises, stems with the Porter algorithm and removes stop-words. Phrasings like "feeling anxious" and "I feel anxiety" then reach the same intent without a trained model.
This early version builds SQL queries by string concatenation and stores passwords as plain text. Today I would use prepared statements, password_hash() and password_verify(), and session hardening, and those are the first changes I would make before any real deployment.