Data Scientist Computer Science, LLM, NLP

RAHUL KHURANA

+1-602-***-**** ********@***.*** linkedin.com/in/rahul-khurana github.com/RahulKhurana-16 EDUCATION

Master of Science of Computer Science May 2025

Arizona State University, United States GPA: 4.0/4.0 B.Tech of Computer Science & Engineering 2019 - 2023 Guru Gobind Singh Indraprastha University, India GPA: 4.0/4.0 EXPERIENCE

Research Assistant Jan 2024 - Present

Geometric Media Lab (GML) Tempe, AZ

• Designed and implemented a scalable, cloud-based Multimodal AI Instructor leveraging RAG techniques and fine-tuned LLMs, improving contextual understanding by 31% and reducing query resolution time by 40%.

• Deployed an end-to-end serverless infrastructure on AWS (API Gateway, Lambda, Step Functions, DynamoDB), optimizing real-time multimodal data processing and cutting instructional support costs by 25%

• Collaborated with cross-functional teams to integrate innovative AI solutions seamlessly into educational operations, enhancing personalized learning for 1,500+ students. Data Analyst July 2022 - May 2023

WSNE Consulting Pvt. Ltd. Delhi, India

• Developed scalable data visualization solutions utilizing Tableau and SSRS, improving real-time insights and reporting efficiency by 20%.

• Implemented CI/CD pipelines for ML models using GitHub Actions, reducing deployment errors by 30%.

• Optimized SQL-based ETL processes and data pipelines to effectively analyze large-scale datasets, significantly boosting operational efficiency.

Machine Learning Engineer July 2021 - Sept 2021

Matrix Geo Solutions Pvt. Ltd. Delhi, India

• Created and optimized high-performance algorithms for ”Geo-Mapper,” achieving 10m accuracy in object localization through advanced image processing and computer vision techniques.

• Enhanced computational efficiency by optimizing YOLO-v4 object detection architecture, reducing error margins by 73% and attaining 89.95% precision in large-scale road damage detection tasks.

• Collaborated across engineering teams to build highly parallelized image processing pipelines, processing over 2000 images/day while reducing processing latency by 50%. SKILLS

Cloud Google Cloud Platform (GCP), Azure, Docker, Serverless Architecture Languages Python, C++, CUDA, SQL

Tools CI/CD (Jenkins, GitHub Actions), Big Data (Spark, Airflow), REST/GraphQL APIs Methods MLOps, ETL/ELT, Microservices

AI/ML Generative AI, RAG, Agentic Frameworks, LangChain, OpenAI, TensorFlow/PyTorch, NLP PROJECTS

DecompDreamer: A Gaussian splatting-based Text-to-3D model Under Review in ICCV 2025 [paper] Oct 2024 - March 2025

• Orchestrated training protocol leveraging VLMs for structured scene decomposition, achieving accurate mod- eling of intricate spatial relationships in complex 3D scenes with up to 11 objects.

• Achieved state-of-the-art performance in compositional 3D asset generation, outperforming existing methods with 54% user preference in evaluations, and demonstrating 75% faster convergence compared to existing techniques.

Leveraging Topological Guidance for Improved Knowledge Distillation Proceedings of the ICML 2024 Workshop GRaM [paper] [code] Jan 2024 - May 2024

• Pioneered a Knowledge Distillation method reducing model size by 30% while increasing accuracy (+1.9%) on benchmark image datasets (CIFAR-10, CINIC-10)

• Leveraged Topological Data Analysis (TDA) to significantly boost robustness against noisy images, achiev- ing 4x faster processing during test-time compared to larger traditional models. SERVICES

NeurIPS 2024 Workshop RBFM Reviewer Sept 2024 - Nov 2024 Provided detailed and constructive feedback

Teaching Assistant at ASU for CSE110: Principal of Programming Course Feb 2024 - Aug 2024 Evaluated over 500 student assignments and assisted students with JAVA programming questions. Machine Learning Lead at Google Developer Student Clubs Sept 2021- Sept 2022 Led a team of 38 members, conducted online & offline technical seminars on LLMs and Computer Vision.

Contact this candidate