Open to opportunities

Aroop Rath

@arooprath

Message

Aspiring AI developer with a focus on machine learning solutions.

India

Message

What I'm looking for

I am looking for opportunities that allow me to innovate in AI and machine learning, with a focus on collaborative environments that foster growth and creativity.

I am a dedicated Computer Science student at Dwarkadas J. Sanghvi College of Engineering, with a strong GPA of 8.50. My passion for artificial intelligence and backend development has driven me to pursue internships that enhance my skills and contribute to innovative projects. During my time at Fixit, I successfully improved model performance by refining accuracy and optimizing backend services, showcasing my ability to leverage Python and machine learning techniques effectively.

My experience as an AI Intern at Cloudscale allowed me to develop a BioNER model that utilized GPU acceleration, significantly reducing inference latency. I am particularly proud of my work on implementing API endpoints that streamlined model deployment, which is crucial for data-driven solutions. Additionally, my research internship at the Indian Institute of Technology, Patna, focused on improving factual accuracy in language models, demonstrating my commitment to advancing AI technologies.

Experience

Work history, roles, and key accomplishments

AI Development and Backend Intern

Fixit

Jan 2025 - Present (5 months)

Enhanced model performance by refining accuracy, improving call transcript extraction and summary generation by 7% through targeted fine-tuning and testing using Python and machine learning techniques. Optimized FastAPI services to ensure smoother backend processing and efficient API performance. Engineered scalable database solutions using MongoDB to facilitate efficient extraction of data from s

Python Machine Learning fastAPI MongoDB API Performance Backend Data Extraction

Artificial Intelligence Intern

Cloudscale

Aug 2024 - Present (10 months)

Developed a BioNER model with CUDA-enabled GPU acceleration, achieving 5% lower inference latency compared to CPU-based execution while leveraging artificial intelligence and natural language processing methods. Refined performance by integrating parallel processing and efficient data pipelines, reducing response time by 17% compared to non-optimized GPU models. Implemented API endpoints to stream

Artificial Intelligence Natural Language Processing CUDA GPU Acceleration API Development

Research Intern

Indian Institute of Technology, Patna

Jun 2024 - Present (1 year)

Developed and implemented hallucination mitigation strategies for LLMs, improving factual accuracy through fine-tuning. Designed evaluation metrics to detect and reduce hallucinations, achieving a 12% decrease in false outputs using benchmark datasets. Conducted extensive research on adversarial attacks and robustness evaluation in LLMs.

LLM Metrics Research