Hello, I'm
Padhmasini A
I'm a
I'm Padhmasini, a passionate and driven pre-final year student pursuing an Integrated Masters in AI and ML at Coimbatore Institute of Technology , Coimbatore . With hands-on experience in LLMs, Machine Learning, and Full Stack Development, I’ve built scalable AI products, including a RAG-based reporting tool during my internship. As Joint Secretary of the FOSS Club, I lead initiatives bridging academic knowledge with real-world impact. Currently, I’m actively seeking full-time opportunities (FTE) where I can contribute to cutting-edge AI/ML teams and continue to grow as a product-focused engineer.
Read More
Worked on building a RAG-based CDP (Carbon Disclosure Project) reporting tool focused on Industry 4.0/5.0 applications.
Responsibilities:
Skilled in ML algorithms (regression, classification, clustering), ensemble methods (XGBoost, Random Forest), and deep learning with CNNs/RNNs using PyTorch and TensorFlow. Experienced in NLP, fine-tuning LLMs, and building RAG-based solutions with Hugging Face and LangChain.
Skilled in statistical analysis, probability, and uncertainty modeling. Proficient in SQL, ER modeling, and database systems. Experienced with Bayesian analysis, heuristic search, and first-order logic for data-driven decision-making.
Skilled in building web pages using HTML, CSS, and JavaScript. Comfortable creating interactive UIs and handling basic backend tasks with Node.js and MongoDB.
This project uses LLMs and semantic search to offer culturally grounded advice based on Thirukkural, the ancient Tamil text of ethics. It provides contextual moral guidance and relevant couplets for real-life dilemmas, combining tradition with AI to inspire thoughtful decision-making. By bridging classical wisdom with modern technology, it empowers users to navigate ethical challenges with clarity, empathy, and purpose.
This project uses Django REST Framework and PostgreSQL to build a robust issue tracking system, streamlining bug reporting and task management for software development teams. It provides structured project and issue organization, enabling efficient collaboration through real-time issue assignment, status updates, and priority handling. By simplifying issue tracking and project monitoring, it enhances productivity, supports agile workflows, and improves development visibility for teams worldwide.
This project uses Reinforcement Learning (RL) to automate financial trading, derivatives pricing, and market-making, simplifying complex decision-making in dynamic markets. It provides intelligent trading strategies and real-time portfolio optimization, enhancing financial analysis and investment performance. By streamlining trading decisions and risk management, it improves financial outcomes and empowers users to develop adaptive, data-driven trading systems.
This project uses CNN to automate cuisine classification and dish recognition, simplifying food image analysis. It provides accurate cuisine predictions and detailed dish metadata, enhancing culinary understanding and exploration. By streamlining cuisine identification and dish recognition, it improves culinary experiences and facilitates recipe discovery for users globally.
This project is a web application built with Python 3.10 and Flask, using machine learning techniques to classify breast cancer tumors as benign or malignant. It preprocesses data, trains an SVM model, and deploys it using Kubernetes. The impact is providing a user-friendly interface for accurate tumor classification, aiding healthcare decisions.