> Incoming SWE at Google. [Summer '25]

> Flomoney Pvt. Ltd. [AI Intern, LawGPT Team, Summer '24]

• Trained and utilised BART transformer model with pipeline abstraction for title and summary generation for chronological analysis and insight extraction of the legal documents.
• Used SVM with CountVectorizers for feature extraction, achieving 99.3% accuracy in document classification.
• Dockerized and packaged it as a containerized application for future testing and deployment on Azure.
• Achieved upto 95%+ accuracy when tested using weighted semantic and syntactic similarity metrics.
• Created a custom library for graphs extraction from documents using combination of LayoutLMv3 and OpenCV.

• Duration: 3 months
• Based out of: Middletown, Delaware, United States · Remote

🡭 Other experience can be found on LinkedIn.