MS Computer Engineering @NYU LLMs, Machine Learning, Neural Networks

JAYRAJ M. PAMNANI

VIEW MY PORTFOLIO

Email me: ********@***.*** LinkedIn: linkedin/in/jayrajpamnani

Call me: +1-646-***-**** Location: New York City, NY 11215 EDUCATION

New York University Master’s in Computer Engineering New York, NY May 2026 Relevant Coursework: Machine Learning, Principles of Database Systems, Applied Matrix Theory, Real Time Embedded Systems, Big Data and Computing Systems & Architecture

(Recipient of a partial scholarship awarded by the university based on my previous academic achievements). Parul University Bachelor’s in Computer Science & Engineering with Specialization in AI Vadodara, India May 2024 Relevant Coursework: Data Structures & Algorithms, Machine Learning, Deep Learning with NLP, Pattern Recognition, Image Processing, GPU Computing, Data Visualization and Data Analytics TECHNICAL SKILLS

Machine Learning & AI: Scikit-learn, TensorFlow, Keras, PyTorch, Random Forest, Decision Trees, Clustering, Regression, NLP (Sentiment Analysis, Text Classification), Deep Learning, Computer Vision Worked with LLMs: Gemini-flash-2.0, Phi-2, CodeLlama-7B, Mistral-7B, Llama2, Zephyr, CodeGen, CodeGemma Big Data & Databases: Hadoop, Apache Spark, Hive, BigQuery, PostgreSQL, MongoDB, SQL, Oracle Data Modeler, MySQL, CosmosDB Programming Languages: Python, R, Java, JavaScript, C++, SQL, MATLAB, TypeScript Data Analysis & Visualization: Pandas, NumPy, Matplotlib, Seaborn, SciPy, Tableau, R (data wrangling, visualization, statistical analysis) Platforms: VSCode, Github, Ollama, HuggingFace, AWS, Google Cloud, Kubernetes, Docker, Azure, Google AI Studio PROJECTS

CommandLineHelper: Natural Language to Bash Command Translator

● Developed an intelligent command-line assistant using Python and CodeLlama-7B that converts natural language instructions into precise bash commands through Retrieval-Augmented Generation (RAG).

● Implemented RAG with ChromaDB and sentence-transformers to enhance command accuracy by learning from historical examples and semantic patterns.

● Built a dual-interface system featuring both CLI and web interfaces, with command history tracking and debug capabilities, making terminal operations more accessible to users of all technical levels. AnimeGAN: Crafting Faces with Deep Learning

● Designed and implemented a Deep Convolutional Generative Adversarial Network (DCGAN) using TensorFlow and Keras to generate realistic anime face images.

● Optimized the generator and discriminator models with advanced techniques to improve the quality and diversity of outputs.

● Visualized training progress and results through Matplotlib to evaluate model performance effectively. Chapter: Secure Library Management System (Project domain given by Professor for Principles of Database Sys. course)

● Designed a robust relational database for a real-world business case using Oracle Data Modeler, translating requirements from the professor into an efficient schema.

● Developed a secure, full-stack web application with Python (Django), HTML, CSS, and JavaScript, implementing features such as SQL injection prevention, deadlock handling, and password encryption.

● Built user authentication with distinct customer and employee logins; employees access a dedicated dashboard that summarizes key business metrics.

HexDrop: Private, encrypted file delivery

● Developed a secure file sharing app using Next.js, TypeScript, and AWS S3 with end-to-end encryption and real-time transfer capabilities.

● Implemented real-time file transfer system with Socket.io and PostgreSQL, featuring progress tracking and download management.

● Built responsive UI with Tailwind CSS supporting drag-and-drop uploads and secure file sharing through unique key generation. PROFESSIONAL EXPERIENCE

Robotskull - Data Scientist Intern; Vadodara, India Dec. 2023 – Mar. 2024

● Streamlined inventory tracking and demand forecasting by leveraging Excel, SQL, and SAS to manage and analyze large datasets, which improved operational efficiency and accuracy in decision-making.

● Reduced surplus stock by 15% by developing predictive models and insightful visualizations using Python and R, which optimized stock levels and minimized overstock costs.

● Enhanced procurement planning efficiency and strategic alignment by collaborating with cross-functional teams to implement data-driven decision-making, resulting in improved synchronization between project requirements and component imports. CERTIFICATIONS

Deep Learning Specialisation by Andrew Ng (5 Course Series)Duration: 3 months Key learnings: Neural Networks Optimisation, Hyperparameter Tuning, CNN, RNN Google Data Analytics Professional Certificate Course (8 Course Series)Duration: 2 months Key learnings: Data Cleaning, Visualization, Statistical Analysis AI-900, SC-900, DP-900 & PL-900 by Microsoft (4 separate courses)Duration: 4 months Key learnings: AI, Data, Automation, Analytics, Security, Identity, Cloud

Contact this candidate