Bhashya Sheth
[email protected] • +1 (571) 394-1663 • https://www.linkedin.com/in/bhashyasheth
TECHNICAL SKILLS
_________________________________________________________________________________________________
Language: Python, C++, JAVA, JavaScript, MySQL
Web Technologies: HTML, CSS, React, NodeJS, ExpressJS, Dash, MongoDB, Django,Flask
ML Technologies: NumPy, Pandas, Matplotllib, Seaborn, Plotly, Scikit-learn, PyTorch, TensorFlow
Cloud: AWS, Kubernetes, Docker, Google Cloud
RELEVANT WORK EXPERIENCE
__________________________________________________________________________________________________
Freelance (Software Developer) January 2022 –June 2023
E-commerce Website (MERN Stack)
● Developed and deployed an e-commerce platform using MongoDB, Express.js, React.js, and Node.js, scaling to
support 500 users and increasing transaction completion rate by 12%.
● Integrated collaborative filtering algorithms for product recommendations, increasing user session duration by
15% and boosting product discovery rate.
● Developed admin dashboard using React and Chart.js, improving product and order management efficiency.
● Optimised front-end load times by 30% using React hooks, lazy loading, and Redux, resulting in smoother user
interactions and a 20% increase in page speed score.
User Behavior Analytics for E-commerce
● Developed a real-time data pipeline using Kafka and Apache Spark to process and analyse user behaviour across
the web, handling over 50,000 events per second.
● Built a hybrid storage solution with Amazon S3 for raw data and Amazon Redshift for aggregated data, optimising
query performance and storage costs.
● Enabled real-time analytics dashboards via AWS QuickSight, delivering key insights on user engagement,
conversion rates, and abandonment patterns to business teams.
Katonic (Software Engineer Intern) June 2021 –August 2021
● Migrated the platform from EJS to a SPA with React.js, reducing maintenance overhead and improving
development speed by 25%.
● Refactored UI components, optimising load time by 20% and increasing session duration by 15%.
● Engineered a full-stack solution with React and TypeScript, enhancing scalability to support 30% more users.
● Collaborated with senior stakeholders, delivering a scalable solution that met business goals ahead of schedule by
3 weeks.
PROJECT WORK
___________________________________________________________________________________________________
Gastrointestinal Disease Prediction System with LLM-Based Analysis and Interactive Dashboard
● Developed a deep learning-based system to classify gastrointestinal diseases using the KVASIR dataset (4,000
annotated images across 8 disease classes), achieving a peak accuracy of 90.1% with the DenseNet121 model.
● Implemented data preprocessing, including image resizing for consistency, RGB format conversion for model
compatibility, and optimised training pipelines using feature extraction and embedding techniques.
● Designed a full-stack application with Flask, Python, and JavaScript, incorporating an interactive dashboard to
display classification results, class-wise performance metrics, and dynamic ROC visualisations.
● Designed and integrated an intelligent chatbot leveraging Meta-Llama-3.1-8B, enabling real-time, context-aware
user interactions and increasing query resolution efficiency
Fake news detection:
● Developed a one-class classification model using a Gated Variational Autoencoder (VAE) trained on real news
data to detect fake news
● Engineered multi-modal features, including Sentence-BERT embeddings, linguistic (POS/NER) metrics, and
domain metadata, to enhance discrimination between real and fake news.
● Compared multiple deep models like CNN, LSTM, BI-LSTM with TF-IDF and Word2Vec embeddings.
● Achieved high accuracy of 94% by leveraging reconstruction error as a decision metric and implemented robust
visualization and evaluation of model performance.
EDUCATION
___________________________________________________________________________________________________
Virginia Tech (Virginia, USA) August 2023 – May 2025
Masters of Science — Computer Science
Relevant Courses: Machine Learning, Deep Learning, NLP, Data Analytics, Cloud Computing,Software Engineering
