π Built a full pipeline for Legal Document Digitization, using YOLOv8 for region detection + OCR + Streamlit UI, check it out here.
π Working on fine-tuning ByT5 on legal datasets to enhance OCR error correction & NER in my legal document application.
π§ Working on a Gen AI solution for Mock Question Series Generation across subjects like Math, Physics, Bio & Chem.
π§° Created a carbon emission calculator using Google Maps API & Flutter + Supabase.
π§ͺ Conducted structured web scraping + keyword extraction for a probiotics firm across 3 verticals.
β¨ Fun fact: I fine-tuned Gemma 2B to drop wisdom with idioms across 72 languages π
π Final year BSc CS (AI & ML) student, Annoyingly curious π§
- OCR and Document AI π
- Deep Learning/ Computer Vision models π€
- Solve some Real-world problems with AI ;)