CareMate
--
Providing Access to mental health experts using AI
Problem Statement
There are doctors and support for any kind of epidemic. Currently, we are at the inflection point for the upcoming epidemic which is the Mental Health epidemic. At present there is a huge shortage of psychologists or mental health experts.
Most kids or people want sessions but the lack of affordable therapists makes it pretty hard. Most people don’t want to invest 1000 RS (12–20 $) per session (and that too for multiple sessions).
Problems with current psychologist Market
So I tried a bunch of psychologists (to understand what techniques are currently used) and my experience was horrible. Most of the psychologists were biased (at the end humans will be biased), their treatment was often time not science-based, and their motive was to make money (that’s fine, but still a motive to make money out of someone else illness isn’t correct)
Most kids or people when they share with their families that they want to get therapy then they are ridiculed, shamed, or not taken very seriously.
So there is a research by health ministry of India that it will going to take 47 years to fulfill the need of mental health experts.
In India, there is lack of mental health infrastructure. Therapist don’t practice evidence based therapy (learning from other cases), backed by research and science.
Most therapist don’t even update themselves after college is finished (noone learn from new data, research, other cases).
Product (Solution)
Currently, I am building 2nd version of the App that uses the LLma2 model (fine-tuned with human-type conversation and psychology Datasets). I tried GPT 3.5 turbo but its costly though it is easy.
AI can truly be revolutionary in solving this problem (Vinod khosla has great articles on this).
The reasons AI will be good fit —
- Updated knowledge
- Learning from other cases (from all over the world)
- No bias (if trained on proper data). Humans have bias, very limited knowledge
- There is a very big problem (I tried 300+ apps, platform in market).
How does this AI work?
Currently because of resource constraints this AI uses LLma2 model (by meta).
But this AI ask lot of question to users and users answers those questions accordingly then AI try to find patterns in it and trained data (which is publicly available and past cases of users).
Users get module. People who are getting therapy for first time don’t know how will all the things going to work.
So at the introduction or first therapy session AI explain all the things about sessions, understand all the things.
One more thing, most people are very bad at explaining things (almost everyone is bad at this). So this tool (have a layer where user’s response got feed and with the help of prompt engineering that answer got enhanced and then sent to AI bot (in order to get most relevant and useful answer).
Reasons why AI won’t be good fit according to people
Empathy — AI can’t understand human, their emotions, problems, etc. This isn’t totally true. There is an app called replika (replika.com). This is AI companion and I tried this app and it felt I am talking to real human (they fine tune gpt 3.5 turbo).
Product Vision In the Long Term
The vision of this product is to provide access to healthcare to people.
WHO recommends regular Medical checkups and expert advice for pregnant women. But in third-world countries middle-class or below-poverty-line category women don’t have access to doctors (and if they have then their visit is pretty limited).
So a proper midwife (AI) will assist pregnant women during their 9 months and prenatal and post-natal period. E.g. At which week what to do, what not to do.
Most first-time parents find it pretty hard to navigate during the care of the first born. So an expert doctor will guide those parents on what to do, when to do.
And many more (Have a conversation with me. I can share some ideas. Without money I don’t know how will I go about building anything.)
Progress So Far
Website Launched :
The app launched : (this screenshot is 2 weeks old) I changed the store listings
I had conversation with 100 people who are mentally ill or seeking therapy from other Platform (in order to understand problem better). And one response was pretty insightful to me.
A boy shared his therapy experience. A rich lady (army husband) called him for therapy at her house. The only thing she did was upsell her. How she again and again gave gurantee to him that he will be good but he need to take 12–15 session with her. Every session will cost him 2k (and that too for 1 hour). So this is pretty interest problem. And many more cases where Patient knew more than psychologist (They just have degree but not intelligence like huberman labs).
Mark Cuban is the founder of the Cost Plus drug company. So I mailed him out of desperation.
Links
website — caremate.netlify.app
App — https://play.google.com/store/apps/details?id=com.caremate.doctor
Some random ideas
Whatsapp Bot
Content Strategy
As per now I am building this alone. So my content strategy is pretty simple. I go about reading 1k-2k articles and whatever I find interesting I post it on twitter(X) and Instagram.
- Teen relationships increases chance of depression (This was a survey published).
- and many more.
How will money going to be used
Currently, I am doing Market Research, setting a Roadmap (to make sure that people really need this product), managing Social media (creating posts and reels), coding (building an App and website), Talking to users, writing content Strategy.
With the money, I can delegate some work to freelancers, and interns.
Invest in infrastructure (like domain name, and APIs.)
I am working on the AI part (it takes time to do stuff).
About me
19-Year-old Full Stack Developer (I learnt coding on my own in high school). I am dropout so I am building this full time.
Building CareMate.
Applied to a bunch of grants, hackathon (I got rejected because I don’t have cofounder).
Some grants, applications asked why did you choosed this particular problem. So I am into blockchain, for last some weeks I am exploring climate change (I read bill gates book on climate change, it was super interesting and beginner friendly).
The reason I chose this problem because I really want to solve some problem and this access to healthcare is really a problem (it may be possible that my solution doesn’t solve problem, but this access is kind of problem). At the same time I have skills and it won’t require much resources as compared to other problems I figured out.
Note
This is just the basic Stuff about this product. If you are interested in this project as an investor or want to join mail me at peckishhuman@gmail.com
Anykind of Grant, loan(without interest), or investment can help alot.
Atleast we can have conversation (for 5–10 minutes).
This is just rough expression (I had done research for like lot of hours, so if you are interested then I can share more about it).
I am constantly working on this product. So if you are not interested as per now then we can chat later about it.
If anyone reading this or checking out product, then this product will look dead to you. But I am currently building this alone, hardly with some money (and that too isn’t much). But this product have some future I am sure.
The reason behind dropping out. My plan was to earn money via doing college with freelancing and interships (I did interships and freelancing in high school). But time has changed alot and things are changing at speed. I mostly learn from reading books (in past 1 month I had gone through 10–12 books, 100k articles, 5–10 podcast). So dropping out was the best option I had figured out. In case nothing worked then I am very good at learning I will learn blockchain (smart contracts or any other skill and then do job). I had a blog on topic blockchain. By the way this is just general thoughts about my product If any investor is reading then let’s have chat. I can share about technology, future of this product.
There are some other blogs also (if you found this blog, idea, product interesting) then you should check those out.
