Hello! Iβm Ambigapathi, a passionate data scientist specializing in Natural Language Processing (NLP) and Generative AI (GenAI). With a strong foundation in machine learning and data analysis, I thrive on transforming unstructured data into actionable insights and building intelligent systems that understand and generate human language.
Oct 2024 - Oct 2024
Data Digger is an intelligent search engine that I developed using LangChain and Streamlit, designed to streamline information retrieval across multiple platforms. With the ability to query resources like Arxiv for academic papers, Wikipedia for verified information, and DuckDuckGo for general web searches, Data Digger offers a user-friendly interface that enhances research capabilities.
In this project, I incorporated advanced language models, enabling users to interact with the search engine through natural language queries. Data Digger is not only a powerful tool for students and researchers but also an essential resource for anyone looking to navigate the vast amount of information available online efficiently.
- Skills: Streamlit, LangChain, Python, API Integration, Natural Language Processing (NLP)
Sep 2024 - Sep 2024
This project leverages machine learning to predict customer churn, aiming to help businesses retain valuable customers and minimize revenue loss. By integrating FastAPI for backend deployment and Streamlit for user interaction, we empower organizations to make informed decisions based on real-time data.
- Skills: Deep Learning, Data Collection, Machine Learning, Pipelines, Problem Solving
Phase 1: Developed a predictive credit risk model using historical loan data and default indicators. Created a credit scorecard categorizing scores into Poor, Average, Good, and Excellent. Built a Streamlit-based user interface for loan officers to input borrower data and get predictions on default probabilities and credit ratings.
Phase 2: Implemented performance monitoring tools and established procedures for Straight Through Processing (STP) after a 2-month trial. Delivered a highly explainable model and comprehensive documentation on performance and maintenance.
- Skills: Core ML, API, Project Implementation, Machine Learning, Problem Solving, Data Visualization
Objective: Developed a predictive model to estimate health insurance premiums based on age, smoking habits, BMI, and medical history, with over 97% accuracy. Built and deployed the model on a cloud platform and created an interactive Streamlit app for underwriters to make real-time predictions.
- Skills: Business Analysis, Random Forest, Machine Learning, Linear Regression, Problem Solving, Exploratory Data Analysis
Developed a Plant Disease Prediction application using TensorFlow and Streamlit to assist farmers in identifying diseases in tomato plants. This project aligns with my educational background in Agriculture, where I gained a solid understanding of agricultural practices and challenges.
Key Features:
-
User-Friendly Interface: Designed an intuitive Streamlit app that allows users to upload images of plant leaves for analysis.
-
Image Processing: Implemented image preprocessing techniques to prepare images for prediction.
-
Real-Time Predictions: Leveraged a TensorFlow model to provide real-time predictions with confidence scores, aiding farmers in making informed decisions.
-
Skills: Agriculture, Deep Learning, Convolutional Neural Networks (CNN), Python, Jupyter
- Python: Data analysis, machine learning, and NLP.
- SQL: Data querying and management.
- Text Preprocessing: Tokenization, stemming, lemmatization.
- Machine Learning Models: Classification, clustering, sentiment analysis.
- Generative Models: Transformer models (GPT, BERT) for text generation and summarization.
- TensorFlow: Building and deploying deep learning models.
- PyTorch: Developing and training neural networks.
- Streamlit: Building interactive applications for NLP and GenAI model deployment.
- Jupyter Notebooks: For documentation and exploratory data analysis.
- Git & GitHub: Version control and collaboration.
- Analytical Thinking: Strong ability to solve complex problems.
- Communication: Proficient in conveying technical concepts to non-technical audiences.
- Collaboration: Experience working in team environments to deliver projects successfully.
- Email: [email protected]
- LinkedIn: Ambigapathi
- GitHub: Ambigapathi-V
- NLP Community: For the continuous support and knowledge sharing.
- Open Source Contributors: For the tools and libraries that make NLP and GenAI projects possible.