Clin-bot: Your personal medical assistant

• 

• 

  Homepage

• 

  Chat UI

• 

  Account creation page

• 

  Login page

Inspiration

Given the underlying framework adopted by UN (SDGs), we aim to propose our own solution to two persistent social, economic and topological problems faced under resourced regions and communities. We propose Clin-bot an easy to access, free, health-advisor which can easily be integrated into any local clinic to provide immediate, precise advice.

What it does

We introduce Clin-bot a clinically aimed chatbot which proposes to reduce trivial clinical load management by providing accurate, expert approved, precise health-care advice, while providing further functionality to both the users and clinics allowing easy access to chat with one another and assess patient treatment via message monitoring, deducing required medical intervention early on.

How we built it

Inorder to create a technologically advance yet interpretable bot, we decided to use rasa (a chatbot framework) working on keyword-based training to provide user defined answers to said queries. This is one of the many features of Clin-bot as it allows clinics to change and redefine queries and belonging responses with ease making for seamless user experience. Front end code consists of basic html, css and minor js to handle user notifications. Front-end to backend communication is facilitated via flasks's dynamic webpage loading allowing for easier user interactions and hence a higher user satisfaction. Finally for prototyping we utilised basic sql express tables to store data while protecting vital info i.e. passwords using python's bycrpt library. We also facilitated sessions and data retrival from previous chats providing further clinical contexts as and where required.

Challenges we ran into

We had multiple obstacles during this project a few being, lack of reliability and data transparency prompting the switch to yml and sql respectively providing paired, learned responses and direct access to user and chat infos, during prototyping. These concerns are being handled in the future by practitioner defined queries and responses with encrypted user info. while storage.

Accomplishments that we're proud of

Initial tests on small test groups and user interactions provided largely positive feedback on the novelty and need of such a tool in the macro-scale Much praise was given to the bot's capacity to handle more than 6000+ queries even at initial testing.

What we learned

This project allowed us to self-learn Mutiple vital skills such as sql, flask interactions, yml, data handling and encryption-based privacy while working on this project. We also learned multiple micro-skills such as task delegation and most essential careful time-utilisation.

What's next for Clin-bot: Your personal medical assistant

Moving forward we want to pilot test this bot in a real-world scenario i.e. clinic gaining insight into practicality and performance while also understanding any concerns presented by practitioners or users thereof.

Built With

• css
• flask
• html
• javascript
• python
• rasa
• sqlexpress
• yml