The problem PriJour solves

It addresses the issue of Mental Health. Here, emphasis is more on early evidence-based non-pharmacological intervention. We believe that most relief for any mental health patient comes when they express themselves.

Our web app is like an emotional spa or a free talk therapist where the users can spill all of their feelings, without anyone judging them, in the form of a private journal. Although journaling has its own merits but to top it all, our web app gives users a personalised emotional spa experience - where the massage is positivity, the scrub to remove dead cells is powerful words and in the end we shower them with positive affirmation!

Our web app allows users to express their day by typing their feelings over a digital canvas which in turn attempts to analyse their sentiments using an ML model to give them what they really need. The data collected over a certain period is rendered in the form of a graph and helps in determining if the person needs psychiatric help, and the analysis can be shared with a doctor. A special Questionnaire is also designed in such a way that it analyses the user's psychological state and determines his/her happiness index.

Not only this, our website also includes a trained bot - “Mamta” to help users to feel positive and happy just how everyone feels on the lap of their mother. This bot keeps a user engaged through interactive questions.

The insights from the user’s journal and his response to those simple set of questions generates a personalised feed of quotes, book suggestions, blog articles, videos and song therapy which works wonders as instead of scrolling those endless twitter, facebook or instagram feeds generally do more bad than good, our web app, Prijour, provides highly selected and limited feeds which gives the user some time for self-introspection. These feeds are conceptualised on the strong scientific Cognitive Behavioural Therapy(CBT) principles. This positive, undebatable content lushes them with life.

Challenges we ran into

Time Limit : Finishing the project in the given amount of time was difficult.

Training Bot : Training the bot for various user inputs was a tedious task especially as they rely on self-crafted rules.

Prediction Model : Sentiment analyser works better if the model is trained for well-structured and relevant data which was hardly available in the desired format. Shortage of high-end computing capabilities (GPUs) led to selection of a simpler model for sentiment prediction.

The web app processes a lot of data input by the user, analysing it using ML models and then rendering content based on what the user is feeling was a tough job.

Also designing the databases’ schema was within itself another project