I'm Kshitij Gupta
Machine Learning Engineer
Specializing in Large Language Models, Deep Learning, and Production ML Systems
I'm a Machine Learning Engineer at Chubb Business Services India, where I develop and deploy cutting-edge ML pipelines and AI solutions. I completed my undergraduate from BITS Pilani, Pilani Campus majoring in Electrical and Electronics Engineering.
With expertise in Large Language Models, NLP, and scalable ML systems, I've contributed to reducing model inference latency by 40% and improving throughput by 50% through advanced optimization techniques. My work spans from fine-tuning LLaMA-based models to developing production-grade REST APIs and implementing robust CI/CD processes for ML deployments.
I'm passionate about leveraging AI to solve real-world problems and have published research in top-tier conferences including ACL, AACL-IJCNLP, and ACIIDS. My recent projects include AI-driven recipe generation and resume parsing systems that demonstrate the practical applications of modern AI technologies.
• Developed and deployed an end-to-end ML pipeline that fine-tuned a LLaMA-based large language model on domain-specific data, resulting in 25% improvement in inference accuracy and a 15% reduction in model drift
• Engineered a high-performance inference service using vLLM, achieving a 40% reduction in latency and a 50% increase in throughput by leveraging dynamic batching and GPU memory optimizations
• Developed and deployed a Spring Boot-based RESTful API for data mastering, enrichment, and certification, leveraging Maven for streamlined dependency management and advanced caching to improve performance
• Integrated robust CI/CD processes, monitoring, and automated scaling strategies to ensure continuous model improvement and reliable production deployments across cloud-based environments
• Developed a machine learning language model tailored for English-Malay code-switched data, achieving a 20% improvement in accuracy over baseline models by implementing advanced statistical and neural augmentation techniques
• Integrated linguistically informed algorithms—including part-of-speech tagging and grammatical coherence—to enhance multilingual NLP robustness and advance the state-of-the-art in code-switching language processing
Built a web app that keeps track of inventory of Army Camp in a Digital Ledger through an intuitive UI
Assemble AI is my open-source initiative dedicated to leveraging AI for real-world problems. Through this platform, I create innovative LLM-powered tools that demonstrate the practical applications of modern AI technologies, from personalized content generation to intelligent data processing.
SmartChef is an intelligent recipe generation system that utilizes GPT and advanced NLP techniques to craft custom recipes based on available ingredients, cuisine preferences, dietary restrictions, and nutritional needs.
Key Features:
Technologies: GPT, OpenAI API, Python, NLP, Docker, AWS
Insight is an advanced resume parsing system that leverages GPT technology to extract structured insights from diverse resume formats, significantly improving recruitment efficiency and candidate recommendations.
Key Achievements:
Technologies: GPT, OpenAI API, Python, Multi-modal AI, Data Processing
Interested in collaborating on LLM tools or learning more about these projects?
Python, Java, C, C++, C#, JavaScript
PyTorch, TensorFlow, Transformers, Hugging Face, vLLM, Keras
AWS, Azure, Docker, Kubernetes, Git, GitHub
Flask, Spring Boot, Databricks, Maven, LaTeX
SQL, RDS
Large Language Models, NLP, Computer Vision, MLOps