This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle
What I Built
I built StudyBuddy AI - a local-first AI tool for students like me who have 100+ pages notes before exams.
It takes long lecture notes and gives:
- Short summary in 30 seconds
- 5 important points
- 5 quiz questions for revision
As an AIML student in Ramagundam, I made this because I spend 15 hours reading before exams. My friend Priya tested it and said "idi chala useful, Rs.100 pay chestha!"
Demo
I built it locally using Ollama. It runs offline on my laptop.
Video demo: I will upload my screen recording here (For now, local demo works - 20 pages ML notes -> 1 page summary in 30 sec)
Code
GitHub: https://github.com/sanjana-dev/studybuddy-ai
Code is simple:
- Uses Google Gemma 2B model via Ollama
- 20 lines Python: client.chat(model='gemma2:2b', messages=[...])
- No cloud API needed
Who Would Pay For It?
- Students: Rs.99/month to save 5 hours before exams
- Coaching centers near my college: Rs.500/month to give summaries to 40 students
- My juniors who have huge first-year syllabus
With 100 students, that's Rs.9,900/month side hustle!
Why Does Open Innovation Matter?
This side hustle ONLY works because of open-source AI:
Cost = Rs.0 vs Rs.1500: If I used ChatGPT API, 1 student costs me $15/month, but I charge only Rs.99. I would lose money! Gemma + Ollama is FREE and runs on student's laptop.
Privacy: College internal notes are private. Sending to US servers is not allowed. Local Gemma keeps data 100% private.
Offline works: In my hostel, WiFi is slow. Open model works without internet. Closed API fails.
Custom for Telugu: My notes are Telugu + English mix. I can fine-tune Gemma for Tanglish later, closed models can't.
Open-source AI made a Rs.99 tool profitable and private. Closed AI makes it impossible!
Top comments (0)