Post Content:
Welcome to my submission for Week 3 of the Hacktoberfest Open-Source AI Challenge! The theme this week is Slow and Steady. This inspired me to tackle a problem for the ultimate lazy person: writing online classified ads.
The Problem
Whether you are trying to sell a used electric scooter or looking to rehome a pet, writing a good online advertisement is tedious. You have to sit there thinking of a catchy title, list out the features in nice bullet points, and try to make it sound professional. It takes way too much time and energy.
The Solution: SlothSell
I built SlothSell to embrace the laziness theme completely. SlothSell is a lightweight web application that takes a very lazy, short input like "Metro Premium plus 3 wheeler Electric Scooty" and instantly turns it into a highly detailed, professional classified ad that is ready to copy and paste.
It generates a catchy title, highlights the best features in bullet points with emojis, and adds a nice structure so you do not have to do any heavy lifting.
How I Built It
I put this together on my Ubuntu setup using Python. Here is the tech stack I used:
UI Framework: Gradio. This made building the web interface super fast and clean.
AI Engine: I used the open-weight Llama 3 model.
Inference: Instead of down a massive model locally, I connected to the model using the Groq API to get lightning-fast responses without slowing down my machine.
Why Open Innovation Matters
Using open innovation like Llama 3 is a massive advantage for projects like this. When we rely on open-weight models, we are not locked into a single proprietary ecosystem or forced to pay expensive subscription fees just to build simple automation tools. By pairing an open model with a lightning-fast API like Groq, developers can create incredibly lightweight and powerful applications that run smoothly without needing a supercomputer or a massive local GPU. Open innovation keeps the power and flexibility in the hands of the builders. Show Your Work
Here is the interface in action!
You can check out the full code in my GitHub repository here:https://github.com/AbdullahRaoDev/SlothSell.git