Hi there! I'm Jaydeepsinh Rajput, a Machine Learning Engineer with a drive to contribute to open-source projects. I'm currently working as a Data Science Developer 2 at Oracle.
- 🔭 I’m currently working as a Data Science Developer 2 at
.
- 🌱 Delving into Natural Language Processing, Generative AI and Distributed Training at Scale.
- 😄 Pronouns: He/His/Him.
- ⚡ Fun fact: I enjoy Exploring trails on my bike, conquering the pool laps, Always up for trying a new restaurant or exotic dish and Cheering on my favorite teams, playing for fun. Sometimes, Schooling myself on the nuances of stock trading and investment strategies.
- 📫 How to reach me:
📃 Resume
🐍 Python, C++ (Beginner)
🤖 Machine Learning, Natural Language Processing, Data Science, Deep Learning, Stable Diffusion, Language Modeling, Generative AI, Artificial Intelligence, Large Language Models, LLM-Finetuning
🛠️ PyTorch, Transformers, Datasets, Accelerate, Diffusers, PEFT, Tokenizer, Timm, Langchain, Fastai, TensorFlow, Keras, Scikit-Learn, SciPy, Matplotlib, Pandas, NumPy, CUDA
☁️ Oracle Cloud, Azure Cloud, AWS, Google Cloud, OpenStack
🚀 Microservices, Docker, Kubernetes, GitLab CI, Terraform
🌐 Django, Flask, Streamlit, Gradio
🔗 Git, GitHub, GitLab, Confluence, Jira
🛠️ Postman, Rancher
👨💻 Company: Oracle
📆 Dates: October 2021 - Present
📍 Location: Bengalore, India
Achievements/Tasks:
- 🚀 Building a Data Science and Analytics platform offering a pre-configured multi-tenant environment to accelerate (by approximately 20%) Data Science, Machine Learning, Deep Learning, and Generative AI projects for Oracle Teams and their clients using Oracle Cloud services.
- 🔗 Integrated OCI (Oracle Cloud Infrastructure) services such as OCI Data Science, Generative AI, Generative AI Agents, and Digital Assistants with our Data Science and Analytics platform.
- 🎯 Maintaining Gitlab, custom Python Package PyPi registry and Conda environments for diverse Machine Learning tasks.
👨💻 Company: Tata Consultancy Services
📆 Dates: July 2017 - October 2021
📍 Location: Hyderabad, India
Achievements/Tasks:
- ⏳ Decreased data annotation time by approximately 50% by introducing a tool for annotating data for NLP/NLU tasks, specially for Named Entity Recognition tasks.
- 🎯 Acquired a customer by building a Proof of Concept (PoC) for Anomaly Detection from images using a deep learning model with 92% accuracy.
- 🚀 Built vehicle detector and tracker using YOLO for Automated Parking Allocation System.
📖 Master of Computer Application
📆 2015 - 2019
📍 Gujarat Technological University - Ahmedabad, India
👨💻 Company: Oracle
- Developed a Proof of Concept (PoC) for a Retrieval Augmented Generation (RAG) application using Oracle Cloud user guide, incorporating Hugging Face pipeline and FAISS database to store embedding vectors.
- Integrated Langchain for chat functionality to enhance the RAG application's capabilities.
- Utilized LLAMA 2 framework to build and optimize the Oracle Cloud-based RAG application for efficient performance.
👨💻 Company: Personal Project - minimalai
- Developed 'MinimalAI', a custom Deep Learning framework using Python, from scratch, enabling flexible adaptation for various use cases.
- Developed comprehensive notebooks encompassing diverse subjects, exercises, and practical applications following Jeremy Howard's Deep Learning Foundations to Stable Diffusion course.
👨💻 Company: Tata Consultancy Services
- Developed a Named Entity Recogniser (NER) and Relation Extractor (RE) model achieving an accuracy rate of approximately 81% for Nokia's technical documents to enhance information retrieval.
- Developed a table parser utilizing Python to extract Named Entities and Relations from document tables, achieving a remarkable accuracy rate of approximately 92%.
- Automated Openstack cloud VM backup & restore and MongoDB dump process, reducing manual backup time by 70%.
👨💻 Company: Tata Consultancy Services
- Developed and implemented a vehicle detection system utilizing the YOLO algorithm within the Darknet architecture, achieving an accuracy of approximately 84%.
- Integrated tracking and counting functionality for vehicles by utilizing Dlib and OpenCV libraries.
👨💻 Company: Tata Consultancy Services
- Developed a custom Convolutional Neural Network (CNN) architecture tailored to the specific characteristics of circuit images, achieving a high accuracy of approximately 96%.
- Utilized Python and Keras to preprocess circuit images and train the CNN model for anomaly detection.
🚀 Create an intelligent document processing solution with Azure AI Document Intelligence (Microsoft, 2023)
👁️ Build an Azure AI Vision solution (Microsoft, 2023)
📚 Build a natural language processing solution with Azure AI Language (Microsoft, 2023)
🌐 Finetuning Large Language Models (DeepLearning.AI, 2023)
🔍 How Diffusion Models Work (Deep Learning.AI, 2023)
💡 Deep Learning Foundations to Stable Diffusion (Fast AI, 2023)
🤖 Practical Deep Learning for Coders Part 1 (Fast AI, 2023)
🌩️ Oracle Cloud Data Science Professional (Oracle Cloud, 2022)
🔮 Oracle Cloud Foundations (Oracle Cloud, 2021)
🛠️ Applied Machine Learning Course (Applied AI Course, 2019-2020)
🧠 Turing Machine - Digital Machine Learning Engineer (Tata Consultancy Services, 2018)
📧 Email: [email protected]
💼 LinkedIn: jaydeepsinh-rajput