cv

Machine Learning Engineer (Project Associate)

Indian Institute of Science, Bangalore (AIREX LAB)

Developing advanced machine learning models and microservices with a focus on high-performance AI solutions

• Developed ML models for radar intercept point classification, improving accuracy by 25%
• Created microservices using Golang with Apache Kafka for scalable data streaming
• Built high-throughput LLM REST API services using FastAPI, Kubernetes, and Apache Kafka

Research Intern

National Institute of Technology, Kurukshetra

Focused on medical image segmentation and robotics solutions for autonomous systems.

• Conducted segmentation of renal blood vessels using the Deeplabv3+ model, achieving an ROC of 72%, 10% higher than the previous benchmark
• Designed and developed an autonomous Rubik's Cube solver using advanced robotics and algorithmic optimizations, including DIP techniques and the CFOP algorithm
• Curated and annotated a Rubik's Cube dataset for face detection, achieving an accuracy of 98%

Research Intern

University of Central Florida

Leveraged advanced ML techniques to enhance performance in pose estimation and chemistry-based tasks using LLMs and VLMs.

• Applied Vision Transformers and CNNs for pose estimations on NTU-120 Skeleton Points Dataset, improving accuracy by 10% to achieve 95%
• Curated datasets for chemistry-based task predictions involving finetuning tasks for Vision Language Models like Llama Adapter, BLIP, and CogVLM
• Leveraged LLM and VLM methodologies (e.g., BLIP2, Llava 1.5) to surpass benchmarks in chemistry-based tasks by over 20% (Work Submitted at CVPR 2025)

Software Intern

Microsoft Team - Cloud & AI

Collaborated on cloud and AI solutions, implementing integration and deployment improvements

• Implemented integration test cases and bug fixing
• Integrated PowerBI template app solutions for Azure
• Streamlined deployment processes for Azure Marketplace and Solution Center

B.Tech

National Institute of Technology, Kurukshetra

Electrical Engineering

• Data Structures and Algorithms
• Operating Systems
• Soft Computing

JEE Mains Rank 15244

National Testing Agency

Secured All India Rank of 15244 (Top 1.3 Percentile) among 1.1 Million Students

Microsoft Engage Mentee

Microsoft

Selected out of 20K applicants

Smart India Hackathon National Finalist

Smart India Hackathon

Top 5 teams selected out of 500+ Teams

MoP2-Bench: Large Vision Language Models for Molecular Property Prediction

CVPR 2025

Under review publication exploring vision language models for molecular property prediction

LesionLensPro: Skin Cancer Detection using an Ensemble Approach

8th International Conference on Micro-Electronics and Telecommunication Engineering

Accepted paper on skin cancer detection using machine learning techniques

Into The Future

Multifaceted project incorporating face recognition, attendance tracking, and gesture-controlled interfaces

• Implemented secure face recognition using triplet loss and siamese network
• Created automated attendance tracking system
• Developed gesture-controlled drawing and reading board

Even More Fruits

Machine learning fruit classification web application

• Developed ML model using MobileNetV2 with 95% accuracy
• Created full-stack web application with React.js and Flask
• Deployed on Microsoft Azure