Arun Sehgal

About Arun Sehgal

I specialize in deploying open-source LLM models, including Meta LLaMA 3.1 with 405 billion parameters, using LangChain. My expertise includes deploying LLMs on local machines and fine-tuning them for specific use cases.In addition to my work with LLMs, I have extensive experience with cutting-edge technologies such as TensorFlow, Keras, PyTorch, OpenCV, and the broader Python Data Science stack. I excel in implementing Deep Learning Architectures from research papers, computer vision, reinforcement learning, cross-validation, classification, and recommendation systems.I am also skilled in deploying Deep Learning models to production environments, leveraging platforms like TensorFlow Serving and browser-based models. Creating insightful data visualizations with Tableau & Grafana is another area of my expertise.As an AWS Solutions Architect (Associate), I have hands-on experience with EC2 (especially p2.xlarge instances - GPU enabled), S3, Elastic Beanstalk, Lambda, and Amazon Machine Learning/SageMaker. My skills also extend to the Google Cloud Platform, where I focus on deploying and optimizing open-source LLM models for diverse applications.

Experience

1. Data Scientist

   Publicis Sapient

   Jun 2025 — Present · Gurugram, IN

Education

• Punjabi University

  Bachelor of Technology, Computer Engineering

  2013 — 2017

• Birla Institute of Technology and Science, Pilani

  Master of Technology - MTech, Data Science & Engineering