Before I get into the details, I just wanted to mention that out of the other 5 ideas I’ve done, this has been by far the fastest one that I’ve gone from idea to launch. It took me about 3 days to build and launch this. So I think all those reps I’ve been doing (and a library of code snippets I’ve accumulated) are starting to pay off! Alright, let’s get into it!
When I spoke to friends and family that owned pets, they mention that because of the pet-boom from the pandemic, it has been significantly harder to get an appointment with a vet. This makes sense, there is a limited supply of vets and a lot more pets! Since it takes a while for someone to become a vet, understandably, this is going to be an issue for the forseeable future.
As anyone involved with technology has seen recently, ChatGPT has taken the world by storm. Not just because of the powerful LLM (Large Language Model) behind it, but also due to the simplicity of the interface. It finally delivered what all the chat-bots from a few years ago promised to be. So like many of us, I started looking into it, playing with it, testing out different prompts and after playing with the OpenAI API, I thought this would be a great solution for the problem!
I think Ruru can help people answer the less urgent questions they have about their pets. Whilst making sure that their local vets can handle the more physical nature of their pets issues. I think this is a win-win for everyone involved.
Trivia: Ruru (or morepork) is an owl that is native to New Zealand and Australia. Also, it is the name that I used for a last-mile delivery service that my mate and I tried to push at the start of the pandemic. Since that didn’t pan out, I thought it’d be cool to bring it back for this project.
The pet-industry has blown up due to the pandemic. And although it’s growth has slowed down since, it is still a growing industry. So, I think that there is a lot of room for Ruru to grow in the future.
In terms of competitors, similar to the human healthcare industry, there are tonne of companies providing a “Chat with a Vet” online service. Most of these will connect you to a vet that you can either chat online, speak on the phone or over a video call. Some provide 24/7 service and others provide it after-hours when your local vet would not be open. I have not found any that provided this service through an AI yet, but I’m on the lookout for one.
I don’t think the technology is there yet to provide a full-service chat-bot that can diagnose your pet’s issues. However, I think we’ve gotta start somewhere!
I built the Ruru chatbot on top of OpenAI’s API. I did have to setup some prompts so that the chatbot would stay in the pet-related questions and to make sure that it does not try to answer questions that are outside of its scope. Especially since one of the bigger issues with ChatGPT has been that it can be quite confident in providing a non-truthful answer.
I also built a simple credit system that measures the number of questions that the user sends through and charges them based on this.
For those interested, here is the tech stack:
As mentioned earlier, I build a credit system, so at the moment the price is $1 per 10 questions. I’m not actually sure if this covers my costs of just running it, but I thought it is a pretty good price to get people interested in it. I decided to go with the usage-based pricing, as I don’t think people will use it consistently for X number of times in a month, but rather they will use it when they need it.
To entice new users, I’ve also set it up to provide the first 10 questions for free. Again, I’m not sure if this is a good idea, but I think it will help people get a feel for it and see if it is something they would use.
In saying all this, I’ll monitor the costs and adjust my pricing accordingly. In the future, I could potentially offer a monthly or annual subscription for unlimited usage. But I’ll see how people use it and see if that’s necessary or not.
So far I’ve just posted this on different platforms:
It has been a week since launch, here are the results:
So for the first week, I think this is a good start. I will need to figure out how I might be able to convert people from the free tier to the paid tier.
Alright, what’s next with this. Well, I think this has been a lot fun to work on. I learned a tonne on the AI side of things and starting to grasp that little bit more on how all of this works. So eventually I’d like to move away from using OpenAI’s API and host my own model. Mainly because I want to be able to get it to answer the questions based on data that I know is correct. So will have to find a database of documents around pet care that has been vetted by a vet. If anyone knows where to find this, please let me know!
Other features I think might be useful is providing more context with the AI on the pet. So for example, if you have a dog, it might be useful to know the breed, age, weight, etc. This will help the AI to provide more accurate answers. I think this is something that I’ll have to look into more.
Well, that is it for now. It’s good to be back into building things again. I hope you enjoyed this post and as always, if you have questions, feedback or need some encouragement to launch your own project, please reach out!
On a side note, I’m looking for contract work to help fund these projects until I can get them to a point where they can sustain themselves. You can find me on LinkedIn or email me at firstname.lastname@example.org.