🌏 Mumbai, Maharashtra – 400025 | 📞 +91 836 918 4996 | 📧 [email protected]

I'm a passionate Data Scientist skilled in leveraging data to solve real-world problems. My journey in data science is driven by curiosity and the endless possibilities of data.

🎓 Education:

• Postgraduate Degree in Predictive Analytics from the BSE Institute Ltd
• Bachelor's degree in Physics from the University of Mumbai

🚀 Skills:

• Programming: Python, R, SQL and NoSql
• Data Science: Generative AI, LLM models, Stable diffusion, Vision language models, Machine Learning, Deep Learning, NLP, Computer Vision,
• Tools: Git, GitHub, Docker, PowerBI, Tableau, SQL, AWS/GCP, MongoDB, Postgres, Notion, Slack, Jira,Trello
• Frameworks: TensorFlow, Keras, PyTorch, Numpy, Sci-kit learn, Pandas
• Languages: English, Hindi, Marathi, learning Spanish and German

• Fine-tuning Stable Diffusion Model on a Custom Dataset: Fine-tune a stable diffusion model on a custom Dataset to generate high-quality, detailed Images.
• Vision-Language Models for Textile Design Description: Explored the capabilities of multiple Vision-Language models to generate detailed descriptions for textile design images.
• VisDrone : object detection and classification on drone-captured imagery using YOLO (You Only Look Once) models.
• Cloud Cover Detection in Satellite Images: I developed and trained a U-Net model to accurately detect cloud cover in satellite images.
• Image Generation: Demonstrating advanced image generation techniques using diffusion models and inpainting algorithms
• PowerPoint Presentation Elements Detection: Developed an automated solution using YOLOv8 for detecting elements in PowerPoint presentations.
• Resume Optimization Suite: An advanced tool leveraging language models for interview preparation and resume analysis.
• Movie Recommendation System: Built a system using text vectorization for personalized movie recommendations.
• Web Scraping with Selenium: Automated data extraction from LinkedIn using Python and Selenium.
• Metabolic Syndrome Analysis Using ML and DL: Metabolic Syndrome Analysis: Exploratory Data Analysis and Predictive Modeling.
• Machine Learning for Hotel Reservation Predictions: Predicting Customer Cancellations using Machine Learning Techniques.
• Image Classification Project: Developed a Deep learning model for classifying images

As a Data Scientist at Giant Analytics, I specialize in Generative AI and NLP, leading cutting-edge projects that push the boundaries of AI innovation. My role involves close collaboration with international clients and senior data scientists from the US and Ireland, allowing me to stay at the forefront of technological advancements. Key projects and contributions include:

• Team Leadership: Led a team focused on image generation and vision-language models, ensuring project milestones were met.
• Client Interaction: Facilitated clear communication of project requirements with clients, ensuring alignment with their vision.
• Research and Development: Researched advanced methodologies and architectures to improve model accuracy for image generation.
• Model Training: Trained models on custom datasets using techniques like LORA and full fine-tuning for optimized performance.
• Pipeline Construction: Developed a custom pipeline combining Vision and LLM models for enhanced results.
• Prompt Engineering: Curated a prompt-to-prompt method, increasing user prompt accuracy by 90%.
• Model Deployment: Deployed fine-tuned models on Hugging Face for client testing and optimized them based on feedback.
• AWS Sagemaker: Utilized AWS Sagemaker for production deployment, ensuring scalability and reliability.
• Custom Gradio Development: Built a customized Gradio interface tailored to client needs.
• Documentation and Presentation: Documented project processes and presented workflows to stakeholders, ensuring transparency and understanding.

• Proof of Concept: Created a proof of concept for tattoo image generation, collaborating with a client from Malta.
• Requirement Documentation: Researched and developed a comprehensive requirement document to guide the project.
• Prompt Engineering: Engineered prompts for tattoo images, allowing for easy modifications by artists.
• Model Utilization: Employed SDXL and SD3 models for generating high-quality tattoo designs.

• UI Design: Designed custom UI workflows for image generation, enhancing user experience and functionality.

• Model Creation: Led the development of a CLIP model using PyTorch, tailored to specific recommendation needs.
• Dataset and Architecture: Built a unique dataset and modified the model architecture to improve recommendation precision.
• Model Enhancement: Achieved greater accuracy and relevance in recommendations through continuous model refinement.

• Algorithm Development: Engineered an algorithm to convert unstructured documents into structured formats.
• Pipeline and API Integration: Developed the necessary pipelines and API connections, deployed using Docker.

• Document Conversion: Transformed documents into PowerPoint presentations, integrating FastAPI with AWS services for seamless operation.

• Statistical Methods: Implemented innovative statistical methods and a grid-based framework to enhance recommendation accuracy.
• API and Docker Deployment: Created scalable FastAPI APIs and deployed them using Docker for efficient performance.

• Deployment Optimization: Enhanced deployment efficiency using Docker, AWS, and Nginx.
• Infrastructure Management: Managed Docker images and deployed projects on AWS EC2, Sagemaker, and Bedrock environments.

• Object Detection System: Developed an object detection system, improving accuracy from 62% to 94%.
• Data Extraction: Extracted data from PowerPoint presentations and PDFs using advanced web scraping techniques.

• RAG-based Chatbot Development: Created a RAG-based chatbot for healthcare domain queries in local languages.
• NLP Task Integration: Integrated sophisticated NLP tasks and custom document indexing solutions.
• Deployment: Deployed the system on an EC2 Instance with Nginx proxy server for stakeholder testing.

• Inpainting Research: Investigated stable diffusion inpainting techniques for product photography image generation.
• Client Demonstration: Showcased the potential of AI in image generation to clients, demonstrating its value in various applications.
• LLM Model Experimentation: Explored models like Gemma, Llama, Mistral, and KERAN for diverse use cases, expanding the company's AI capabilities.

• Global Teamwork: Collaborated with colleagues from the Netherlands, Chennai, Ireland, and Singapore, fostering innovation and enhancing project outcomes.

This role has allowed me to hone my skills in a variety of areas, including machine learning, data processing, and cross-functional collaboration while contributing to cutting-edge solutions in data science.

• Web Research: Demonstrated strong research skills with a focus on data accuracy and detail.
• Statistical Techniques: Proficient in various statistical methods for data analytics.
• Cloud Platforms: Practical experience with AWS and GCP services.

• LinkedIn: LinkedIn Profile
• GitHub: GitHub Profile
• Instagram":Instagram Profile

• Astronomy and Stargazing
• Reading Books
• Traveling and Trekking
• Swimming
• Cycling
• Trading

• Data Science with ML, AI, GUI & VUI with Libraries of Python (Udemy)
• NISM-Series-VIII: Equity Derivatives Certification
• Affective Computing NPTL

Feel free to reach out for collaborations or just a chat!

📧 Email: [email protected]

👤 Aditya Ghadge