ThiHacktoberfest Open-Source AI Challenge Week 3: Slow and Steady
What I Built
I built UP Cricket Adda Summarizer - an AI tool that does my lazy work.
Every day in UP, there are 100+ local cricket matches. I used to spend 3-4 hours reading long WhatsApp reports and commentary to write a small summary.
Now my tool does it in 10 seconds. It takes a long cricket match report and makes a short, clean summary in Hindi + English.
Demo: https://rajveer-singh-up.github.io/Up-cricket-summarizer/
View Code: https://github.com/rajveer-singh-up/Up-cricket-summarizer
Who gets the time back?
- Me: I save 3 hours daily.
- Local cricket fans in UP: They get fast updates without reading long articles.
- Small cricket pages: Who don't have a big team.
Why Open-Source AI is Better (Slow and Steady)
I tried closed AI APIs first. Two problems:
- Internet needed - in Karsog internet is slow, API fails.
- Pay every time - 100 summaries = too much cost.
So I used open-source model (BART) that runs on my own laptop.
Benefits:
- Works without internet - data stays private on my machine.
- Set up once, never pay again.
- I can change it for Hindi cricket words like Chakka, Wicket.
This is Slow and Steady - build once, own it forever.
How I Built It
- Frontend: HTML, CSS, JavaScript for fast loading on low-end phones.
- AI Model: facebook/bart-large-cnn from Hugging Face.
- Logic: Long text -> Model -> 3-line summary.
- Hosting: GitHub Pages.
I saved my agent session with DevRelay while building.
What's Next?
Add voice output so farmers can listen to summary while working in field.
Made from Himachal for UP Cricket.s is a submission for the Hacktoberfest Open-Source AI Challenge Week 3: Slow and Steady*
Top comments (0)