Hi there! 👋
I'm Katam Vamsi Krishna, a passionate Machine Learning Engineer. My primary areas of expertise include Machine Learning (ML), Natural Language Processing (NLP), and Large Language Models (LLMs). I am constantly exploring new technologies to solve complex real-world problems and improve user experiences through automation, optimization, and predictive modeling.
- Current Focus: Specializing in Large Language Models (LLMs) and their applications in NLP, text generation, and dialogue systems.
- Experience: Experience: Over 4 years of hands-on experience designing, building, and deploying machine learning solutions. This includes implementing end-to-end ML pipelines, optimizing models for real-time applications, and deploying scalable solutions on cloud platforms. My journey combines professional expertise with practical exposure from independent projects during my master's studies, where I developed applications like sentiment analysis systems, GAN-based image generators, disease detection models, LLMs fine tuning, research on mechanistic Interpretability of LLMs. These experiences have honed my skills in delivering impactful, production-grade solutions tailored to diverse business and research needs.
- Tech Stack: Python, TensorFlow, PyTorch, Scikit-learn, AWS, SQL, Git, Flask and more.
- Passion: I am deeply committed to leveraging advanced AI and machine learning methodologies to develop innovative solutions that address complex challenges and drive meaningful outcomes. My approach is rooted in a strong foundation of research and practical application, ensuring robust and scalable results.
Here are some of the tools and technologies I've worked with:
- Programming Languages: Python, JavaScript, SQL, C, C++, R
- Machine Learning & Data Science: TensorFlow, PyTorch, Keras, Scikit-learn, Pandas, NumPy, SciPy
- Natural Language Processing (NLP): Transformers, BERT, GPT, Named Entity Recognition (NER), Topic Modeling, Sentiment Analysis, Text Summarization
- Cloud Platforms & Deployment: AWS (EC2, S3, Lambda), Google Cloud (Vertex AI, BigQuery), Docker, Kubernetes, MLFlow
- Web Development: React, Flask, FastAPI, Django
- Big Data & Data Engineering: Spark, Hadoop, DVC, Airflow
- Data Visualization & BI Tools: Matplotlib, Seaborn, Plotly, Tableau, Power BI
- Version Control & CI/CD: Git, GitHub Actions, Jenkins, GitLab CI/CD
Below are some of the key projects I’ve worked on:
This project implements a deep learning-based system for the classification of kidney diseases using medical imaging. Built to assist healthcare professionals, the solution leverages convolutional neural networks (CNNs) to analyze and classify images efficiently and accurately.
This research works is of an in-depth exploration of Chain of Thought (CoT) reasoning in Large Language Models (LLMs), focusing on its impact on solving complex reasoning tasks. The project leverages the GSM8k benchmark dataset and includes experiments with zero-shot and few-shot CoT approaches, along with the Majority Vote strategy for output reliability.
A machine learning-driven web application built to predict and assess credit risk for loan applications. Deployed on AWS for scalable, real-time predictions.
An automated ticket routing system utilizing machine learning algorithms to classify and route tickets to appropriate departments without human intervention.
Developed a Deep Convolutional Generative Adversarial Network (DCGAN) to generate realistic handwritten digit images similar to the MNIST dataset. This project demonstrates the use of adversarial training to create high-quality synthetic data.
Built a Long Short-Term Memory (LSTM) network for sentiment analysis on IMDB movie reviews. The model classifies reviews as positive or negative by capturing temporal dependencies in text data.
Implemented an end-to-end pipeline to classify clothing images from the Fashion MNIST dataset using neural networks. The project includes data preprocessing, model training, validation, and evaluation.
Developed a Convolutional Neural Network (CNN) to identify diseases in plant leaves from image data. This solution helps in early disease detection and agricultural management.
Built a feedforward neural network using PyTorch to predict breast cancer based on clinical data. The model assists in early diagnosis and improving healthcare outcomes.
Check out my detailed resume for a deeper look into my professional experience, education, and achievements.
Feel free to reach out to me for collaborations, open-source contributions, or simply to chat about the latest in AI/ML and NLP!
Thank you for visiting my GitHub! I’m always open to learning new things and collaborating on exciting projects. 🚀