SHOWCASING BRILLIANCE
THE PROJECTS THAT DEFINES US!
Polygon
On-Going
Projects
PROJECT ON LLM MODEL DEVELOPMENT FOR COMPETITIVE EXAMINATION
Project Image

Developers: Vatshal Mehta, Himakshi Arora, Khushi Gupta.

Description: This project develops a multilingual, syllabus-based LLM that generates exam-type questions and answers with native-language explanations.

PROJECT ON LLM MODEL DEVELOPMENT FOR EDUCATIONAL COURSES INFORMATION
Project Image

Developers: Kajal Rathod, Sakshi Dhuri, Riya Birnale.

Description: An LLM-powered advisory system that suggests courses, colleges, and bridge programs with filters like location, fees, and eligibility.

PROJECT ON LLM MODEL DEVELOPMENT FOR DIAGNOSTICS CENTERS AND HOSPITALS
Project Image

Developers: Smit Karelia, Molisha Jain, Sayali Jadhav, Vrundha Parekh, Harish Jalani

Description: An AI assistant that gives quick hospital details, diagnostic info, and personalized health scheme recommendations with multilingual support.

Polygon
Completed
Projects
DEVELOPMENT OF AN AI-POWERED QUESTION BANK FOR JEE AND UPSC EXAMINATIONS
Project Image

Developers: Heer Panchal, Kinjal Panchal, Jhalak Dedhia, Adrita Banerjee, Aniruddh Sengupta

Description: This system will help aspirants save time, practice efficiently, and focus on areas of improvement, making preparation smarter, personalized, and more effective compared to conventional methods.

LEVERAGING LATENT DIFFUSION MODELS FOR DERMATOGLYPHICS DATA ANALYSIS
Project Image

Developers: Prasad D. Devkar, Aastha Bhatt, Mahiman Dave

Description: Fingerprint extraction and analysis from hospital forms using SAM + TDA. Minutiae and persistence barcodes enable similarity matching, improving biometric identification for healthcare data.

AUTOMATED ANSWER SHEET EVALUATION AND RANKING SYSTEM
Project Image

Developers: Ashika Jain, Bhargavi Joshi, Sravan Kotta, Namitha Prakash, Aryan Ilake, Shreyas Konidala

Description: Automated answer-sheet evaluation and ranking using Gemini 1.5 Pro for OCR and scoring. Provides consistent, objective results while reducing human error and bias.