Cuddly
One Stop Petcare and Rehome Solution.
The problem Cuddly solves
-
Addressing the Surge in Abandoned Pets Post-Lockdown:
During the past few years, particularly amid the lockdown period, there was a notable spike in pet adoptions. However, with the lifting of restrictions, many individuals find themselves unable to dedicate sufficient time to their pets, leading to an unfortunate increase in abandoned animals roaming the streets. Our solution addresses this pressing issue by providing a dedicated platform for both pet owners and potential adopters to seamlessly communicate and find new homes for pets in need. With features such as direct chat functionality, preference filtering, and AI integration, our platform equips users with all the tools necessary to facilitate the rehoming process effectively. -
Expert Guidance and Support Through Our Trained Chatbot:
Eradicate uncertainties and gain invaluable insights with the assistance of our trained chatbot. Our platform boasts a meticulously trained chatbot specifically designed to offer comprehensive knowledge and assistance on various pet-related matters. Leveraging data from extensive pet care resources, our chatbot is adept at providing tailored solutions to address any concerns or inquiries users may have regarding their pets. -
Enhancing User Experience with Advanced Image Detection:
For users who may lack detailed information about their pets or encounter stray animals, our platform enhances the rehoming experience with advanced image detection capabilities. Seamlessly integrated into our platform, this feature automatically identifies key attributes such as pet type, breed, and color. Whether you're rehoming a pet or encountering a stray animal whose breed is unknown, our image detection functionality ensures a streamlined user experience, facilitating the process of finding suitable homes for pets in need.
Challenges we ran into
-
Vector Embeddings and usage of langchain : At the start, we had a tough time with things like vector embeddings and Langchain. They were new to us and a bit tricky to figure out. Plus, finding good and affordable models for embeddings and making Langchain work smoothly with big language models (LLMs) was tough. Also, getting the right data for our chatbot was a headache. But, we dug deep, did some serious research, and eventually got the hang of it. We found the right models, cracked the Langchain puzzle, and sorted out the chatbot data. Teamwork and persistence paid off!
-
Implementing Caching with redis : As we tried to implement the caching feature the in our platforms to make it more scalable by reducing the database reads, it became more and more challangeing to implement as we were filtering data. We solved this problem by exploring about features of redis including redis json, redis architecture and made the key pairs seperately for individual filters which solved the issue.
-
Implementing Chat using Socket : Adding a chat feature using Socket.IO wasn't easy. Our platform isn't just about chatting, and we don't store loads of user info. So, making real-time chatting work smoothly was a challenge. We had to tweak how our database was set up and find ways to make chatting happen in real-time without slowing everything down. It took some trial and error, but we nailed it in the end by tweaking our database setup and finding smart solutions. Now, chatting on our platform is seamless and fast.
Tracks Applied (1)