Oncology.AI Used AI

This is so beautiful i'm moved, the scrollbar has become global and my project is complete... for now maybe??

feeling like palpatine rn with my unlimited power

I'm definitely harvey specter the way these terms are smooth ashhh
and also i made a custom scrollbar if you look closely yippeee

terms and conditions why is it so hard to make there's so much to type and so much to edit i just don't like legal stuff man, this is all i could come up with for now

managed to shorten a 40+ line of code into just 5 haha, ai really CANNOT code

I'm sure you guys are tired of me refactoring my code but the thing is, there is beauty in simplicity. for example, this sick banner i developed

Encountered the BIGGEST jumpscare of my life and it's due to missing () -_-

Unfortunately writing maintainable code is very painful but here's a funny snippet of me misunderstanding stuff. My time was spent removing the code pertaining to this mishap

Added renal cancer to the list of cancers supported by my app, also updated the readME

I've improved the test prescriptions, only recommending them upon the most stringent of factors. Ngl it feels easier without AI??

Not gonna lie, I'm starting to realize the demerits of using AI. Looked at most of my test prescription file and there were alot of unnecessary checks that i've manually removed. It's very taxing but hey at least i'm learning. Here is a code snippet that i've recently started sheering off.

It was getting confusing to figure out each section was (forgive me for my bad commenting skills) so i used claude to mark each section. Also made a few minor tweaks to the summary component so it's more professional

Need to avoid legal trouble so threw in a bunch of disclaimers and got rid of a few options that shouldn't be given to people trying to sue me and get some money (I'm broke bro)

boom okay i thought making devlogs voided my current ship but HEY guess who's back with a redesigned summary page. Made it simpler because i realized everything is better off minimalist with maximum functionality hahaaaaa

well i wish i avoided the apocalypse (horrific devlog timings) but in the past 3 hours of work I have tested, tested and more tested, i'm happy to announce that my project is done! (for now)

SoM had a few issues with their DB so didnt post a devlog but, a BRILLIANT new landing page for my oncology chatbot haha

Not gonna lie, the summary was a real headache for tabbing between files to pinpoint where the exact user response to certain questions were.

we made it through though, just took a light 2 hours and 45 minutes haha, the summary looks more complete i just need to add one more section and i can redeploy it for the hospital

For unknown reasons, some of my submissions dont take me to the next question. HOWEVER, i've already found the reason. I just forgot to import the ID of the next question smh

With the addition of these many questions for just two cancer types. I've also wanted to update my PDF summary as i believe that's the highlight of this project. It's the proof of the user's completion of the questions and hence, it must be the most refined and polished part of my project

Ran into a few bugs, mainly the fact that some of the flow was looping in on itself. The response objects weren't getting updated leading to my flow verification file believing no input was given yet so it kept resetting back to the same question

Turned all my handler functions modular as it was a hassle to manually update each one. This way the conversationFlow file is the one with the sole power to dictate the order of the chatbot questions.

Upon getting feedback from the hospital (and some of my medico friends). I've decided to expand the number of cancer types my project will cover. For now in the pipeline I'll be focusing on ovarian and gastric cancer

Better front end, improved cancer summary component. Had some weird react rendering issues but have successfully fixed them all

Refactored the code, Spread it out over several components to ensure good practices to implement future updates.

Next step: create an appealing landing page

What works: The main objective of the app (cancer screening) and recommend test prescription works perfectly

PS: A hospital has reached out to me and wants me to implement this for them (paid). Time to kick this project into overdrive then.

Fully setup the app, everything is complete and documented in the github. Picture shows a pdf output which is directly obtainable at the end of the chatbot flow

Developed the chatbot flow, to be continued