cv

Technology Analyst

ICICI Bank

Led migration to Adobe's personalization platform across multiple banking applications, impacting 10M+ customers. Built ML-driven recommendation engines and improved system efficiency by 7.5%.

• Led migration to Adobe's personalization platform across RIB, CIB, NLI, and iMobile, impacting 10M+ customers
• Improved efficiency by 7.5% and reduced rendering latency by 1s through optimized data processing pipelines
• Built ML-driven recommendation engines with logistic regression & real-time event processing using Apache Kafka
• Deployed scalable solutions in Unix/Linux environments with 99.9% uptime and automated CI/CD pipelines
• Mentored junior developers and conducted technical knowledge sharing sessions

Research Assistant

AI-ML-NLP Lab, IIT Patna

Developed EDI-Summ, a multimodal summarization architecture for clinical documents, achieving state-of-the-art performance and publishing at ACL 2024.

• Developed EDI-Summ, a multimodal summarization architecture with contextual + cross-attention
• Achieved 0.81 factual recall and 3.51 clinical score, outperforming GPT-3.5 & GPT-4V
• Enhanced BART pipelines with multimodal fusion for state-of-the-art healthcare summarization
• Conducted experiments in Linux-based distributed research environments

MITACS Globalink Research Intern

Polytechnique de Montréal

Built real-time decision support system for climate and water management, processing 20,000+ datasets across 30 cities and improving forecast accuracy by 25%.

• Built real-time decision support system with React.js, Node.js, and ML models
• Processed 20,000+ climate & water datasets across 30 cities
• Improved forecast accuracy by 25% and planning efficiency by 30%
• Deployed on Linux servers with scalable backend services

Software Developer (OSPP & GSoC)

JoomlaCMS

Developed customizable guided tour system with multi-language support, reducing onboarding drop-off by 20% and mentoring contributors in Google Summer of Code program.

• Developed customizable guided tour system with multi-language support & progress tracking
• Reduced onboarding drop-off by 20%
• Mentored contributors in selective Google Summer of Code program (0.2% acceptance rate)
• Collaborated with global teams via Git, Agile workflows, and CI/CD pipelines

Contributor & Mentor

Open Source Community

Active contributor to open-source projects including JoomlaCMS, PyTorch, and various healthcare AI projects. Mentor students in ML/AI and software development.

• Mentored 10+ junior developers and students in ML/AI and software development
• Active contributor to JoomlaCMS, PyTorch, and various healthcare AI projects
• Speaker at local tech meetups and AI/ML conferences
• Teaching coding to underprivileged students

Master of Science

University of Florida

Computer Science

• Large Language Models(LLMs)
• Natural Language Processing(NLP)
• Mathematics for Intelligent Systems

Bachelor of Technology

Veermata Jijabai Technological Institute

Electronics and Computer Engineering

• Data Structures
• Algorithms
• Database Systems
• Software Engineering
• Computer Networks
• Operating Systems

ACL 2024 Publication

Association for Computational Linguistics

Published 'From Sights to Insights: Towards Summarization of Multimodal Clinical Documents' at ACL 2024 - International Conference on Computational Linguistics

University of Florida Graduate Research Fellowship

University of Florida

Graduate research fellowship for outstanding academic performance and research potential in Computer Science

MITACS Globalink Research Internship

MITACS Canada

Prestigious international research internship program for undergraduate students

Google Summer of Code (GSoC) Participant

Google

Selected participant in Google Summer of Code program with 0.2% acceptance rate

EDI-Summ: Enhanced Multimodal Summarization for Healthcare Applications

EMNLP 2024

Under review at EMNLP 2024 - Conference on Empirical Methods in Natural Language Processing. Enhanced multimodal summarization system for healthcare applications.

From Sights to Insights: Towards Summarization of Multimodal Clinical Documents

ACL 2024

Published at ACL 2024 - International Conference on Computational Linguistics. Developed state-of-the-art multimodal summarization architecture for clinical documents.

EDI-Summ: Multimodal Clinical Document Summarization

Developed state-of-the-art multimodal summarization architecture with contextual + cross-attention mechanisms, achieving 0.81 factual recall and 3.51 clinical score, outperforming GPT-3.5 & GPT-4V

• Multimodal fusion architecture
• Contextual and cross-attention mechanisms
• Published at ACL 2024
• Outperformed GPT-3.5 & GPT-4V

Gestural AI – Real-Time ASL Interpreter

Built real-time American Sign Language recognition system with 94% accuracy on 20,000+ videos & 166,000 images using Python, OpenCV, MediaPipe, and TensorFlow

• 94% accuracy on 20,000+ videos
• Real-time processing
• Computer vision and gesture recognition
• Deep learning implementation

Climate Decision Support System

Designed real-time data analysis platform processing 20,000+ climate & water datasets across 30 cities, improving forecast accuracy by 25% using React.js, Node.js, and ML models

• 20,000+ datasets processed
• 30 cities coverage
• 25% accuracy improvement
• Real-time data analysis

JoomlaCMS Guided Tour System

Developed modular onboarding framework with multi-language support integrated into JoomlaCMS core, reducing user drop-off by 20% and improving user experience

• 20% reduction in user drop-off
• Multi-language support
• Modular architecture
• GSoC and OSPP project