cv

Research Assistant

National Institute of Technology, Rourkela

Worked under Prof. Tapas Kumar Mishra on improving neural machine translation for low-resource Indic languages.

• Integrated POS, NER, and lemmatization into Transformer models.
• Used OpenNMT with Samanantar dataset to achieve SOTA performance.
• Contributed to academic publications and internal seminars.

Software Engineer Intern (QDSP Team)

Qualcomm

Enhanced power measurement and inference efficiency for machine learning models on Qualcomm SNPE and QNN SDKs.

• Reduced energy consumption by 30% for ML models.
• Implemented visualization for Inference/watt metrics, increasing analysis efficiency by 40%.
• Improved system reliability by 25% and reduced testing time by 30%.
• Contributed to documentation and cross-team integration.

Instructor

Chittagong Junior Coders

Volunteered to teach programming to school students (Class 2 to 10) in Bangladesh.

• Taught C programming to young learners.
• Helped build early coding skills in underserved regions.
• Collaborated with local Facebook groups to extend impact.

B.Tech

National Institute of Technology, Rourkela

Computer Science and Engineering

• Data Structures and Algorithms
• Machine Learning
• Deep Learning
• Computer Networks
• Database Systems
• Distributed Systems
• Quantum Computing

Winner – IIT Bombay Techfest Hackathon

IIT Bombay

Won first place for developing an innovative AI-driven research prototype in a competitive national event.

Winner – NIT Rourkela Innovation E-Cell Hackathon

NIT Rourkela

Secured first prize in a student entrepreneurship challenge by showcasing a health-tech solution.

Startup Funding – Firefly Startup Incubation Center

NIT Warangal

Received funding and mentorship for developing an Electric Vehicle ecosystem app through Firefly Incubation.

Merit Scholarship – SSC Examination

Government of Bangladesh

Awarded for academic excellence in high school by the national education board.

Study in India Government Scholarship

Ministry of External Affairs, India

Received full financial support for undergraduate studies at NIT Rourkela.

License Plate Recognition System

Developed an image processing pipeline to detect and recognize license plates using CCA, projection, and supervised ML.

• CCA segmentation
• SVC-based character recognition
• 4-fold validation

Traffic Sign Classification with CNN

Built a CNN to classify traffic signs using the German Traffic Sign dataset with 99.33% accuracy.

• Data augmentation
• Y channel processing
• Early stopping & regularization

Custom Image Processing Filters

Implemented blur, edge detection, and grayscale filters for bitmap images using multithreaded C.

• C with multithreading
• Edge enhancement
• Bitmap pixel manipulation

Project Ebony – Webwiz Club

Developed a web portal for learning resources using React and Node.js with CMS and advanced filtering.

• CMS implementation
• Search + filter UI
• Responsive design