Machine Learning Engineer | MLOps Enthusiast
I am Pramit De, a final-year Computer Science student with hands-on experience in Data Analysis, Machine Learning, and deploying AI applications through personal projects and internships.
Interest: My journey into data science started with a fascinating example of EDA revealing hidden patterns, like identifying a hurricane from New York taxi data. This ignited my curiosity, and since then, I’ve been refining my skills through practical projects and internships. I am now focused on solving complex, business-oriented challenges using data science.
- Goal: Develop a lightweight deep learning model for food ingredient classification.
- Why: Automate ingredient recognition for recipes and inventory management in the food industry.
- Outcome: Built a MobileNetV2-based CNN with transfer learning, achieving 93.14% test accuracy with hyperparameter optimization using Keras Tuner.
- GitHub Repository
- Goal: Create an interactive web app for generating high-quality comprehension exercises.
- Why: Enable educators and learners to create customizable comprehension exercises easily.
- Outcome: Developed a Streamlit app using Google GEMINI 1.5 Pro API, reducing errors with a review-generation chain and deployed on AWS EC2 for a fast, responsive user experience.
- GitHub Repository
- Goal: Design a chatbot to recommend recipes based on ingredients and preferences.
- Why: Provide personalized cooking assistance for both home cooks and professional chefs.
- Outcome: Developed a Retrieval-Augmented Generation (RAG) model using cookbooks and Wikipedia data, deployed as a FastAPI app on AWS EC2 with optimizations for faster response times and better memory management.
- GitHub Repository
-
Programming Languages & Tools:
-
Python Libraries:
-
Databases:
-
Core Concepts:
Data Wrangling ,EDA, Statistical Modeling, Machine Learning, Deep Learning
- Business-oriented
- Team player
- Problem-solving
- Strong communication skills
I’m currently expanding my knowledge in:
- MLOps: Streamlining model deployment, monitoring, and CI/CD pipelines to enhance model lifecycle management and operational efficiency.
- Big Data Machine Learning: Utilizing tools like Hadoop and Spark to apply machine learning algorithms to large-scale datasets for distributed computing.
Feel free to connect with me via:
- Email: [email protected]
- LinkedIn: de-pramit
- GitHub: Pramit726
Let's collaborate on data science or machine learning projects.