Shruti B

AI Engineer at ARMA AI Labs INC

Working on audio and video model pipelines using a large language model trained on billions of parameters. Fine-tuning the model with LORA and PEFT techniques to incorporate an emotional layer evaluated across various accents. Leveraging CUDA GPU acceleration with PyTorch to efficiently run and test large-scale models.

Data Science Lab Researcher at Multi-scale Medical Robotics Lab at UMassD

Designed computer vision models using Faster R-CNN architecture to detect tumors and perform breast cancer segmentation on DICOM mammogram images. Extracted regions of interest using bounding boxes from annotation files and extended the project by training multimodal fusion algorithms across multiple image modalities.

Data Analytics Intern at Oceantic Network

Performed data cleaning, manipulation, and visualization to analyze past market trends and business insights for key performance indicators such as job growth in skilled and unskilled laborers. Conducted time series forecasting using ARIMA models to predict a 7% increase in offshore wind job opportunities by 2025.

Software Developer Trainee/Intern at Psiog Digital

Completed a 5-month software development and data analytics training encompassing real-world data team environments. Worked on web development and business intelligence projects, created interactive Power BI dashboards, developed front-end components using React, and performed database operations with MySQL. Explored cloud development using Microsoft Azure.

Data Analytics Intern at Rela Multi-specialty Hospital

Designed an interactive dashboard for hospital call center data to provide actionable business insights involving call volume trends, average handling time, and appointment availability. Created visualizations in Tableau and used R for statistical modeling to improve operational efficiency across customer support.

Machine Learning Intern at Computational Lab at SASTRA Deemed University

Developed and trained deep learning models (CNN, ResNet50, Faster R-CNN) for multi-class breast cancer stage classification and tumor detection on mammograms with high accuracy. Optimized training on UMassD supercomputer to reduce training time. Built Random Forest models for compound inhibitor screening, improving screening efficiency by 90 times.

AI Engineer at ARMA AI Labs INC

Currently working remotely on developing and fine-tuning large language models with billions of parameters, focusing on integrating emotional layers evaluated across various accents. Leveraging CUDA GPU acceleration with PyTorch to efficiently train and test large-scale AI models for audio and video pipelines.

Data Science Lab Researcher at Multi-scale Medical Robotics Lab at UMassD

Designed and implemented computer vision models using Faster R-CNN architecture to detect tumors and perform breast cancer segmentation on DICOM mammogram images. Developed multimodal fusion algorithms incorporating multiple imaging modalities to improve diagnostic accuracy.

Data Analytics Intern at Oceantic Network

Performed data cleaning, manipulation, and visualization to analyze employment trends in the offshore wind industry. Conducted time series forecasting using ARIMA models predicting job opportunity growth influenced by seasonal trends and government policies.

Software Developer Trainee/Intern at Psiog Digital

Completed a 5-month software development and data analytics training simulating real-world project workflows. Contributed to front-end development using React, database operations with MySQL, and gained experience in C#, Angular, and cloud services with Microsoft Azure.

Data Analytics Intern at Rela Multi-specialty Hospital

Developed interactive dashboards based on call center data to monitor KPIs such as call volume trends and appointment slot availability. Used Tableau and R for visualization and statistical modeling to enhance operational efficiency and data-driven decision making.

Machine Learning Intern at Computational Lab at SASTRA Deemed University

Developed deep learning models including CNN, ResNet50, and Faster R-CNN for breast cancer stage classification and tumor detection on mammogram DICOM images, achieving high accuracy with optimized training. Also created Random Forest classifiers for screening chemical inhibitors, significantly improving screening efficiency.

AI Engineer Intern at ARMA AI Labs INC

Working on an audio model pipeline using a large-language model trained on billions of parameters. Fine-tuning with LoRA and PEFT to train a TTS pipeline with an emotional layer evaluated across accents. Leveraging CUDA GPU acceleration with PyTorch to test large-scale models.

Data Science Lab Researcher at Multi-scale Medical Robotics Lab at UMassD

Designed a computer vision model with Faster R-CNN to detect tumors and perform breast cancer segmentation on DICOM MG modality images. Extracted ROIs using bounding boxes from annotations. Extended project to multimodal fusion across multiple imaging modalities.

Data Analytics Intern at Oceantic Network

Performed data cleaning, manipulation and analysis; visualized trends to generate business insights for KPIs such as job-number trends, market trends, and skilled/unskilled labor dynamics. Conducted time series forecasting on offshore wind employment using ARIMA; predicted a 7% increase by 2025; produced stakeholder documentation and BI dashboards in Tableau.

Software Developer Trainee/Intern at Psiog Digital

Developed software components in C# and C++, explored multithreading and distributed systems in Azure; built front-end components with React and performed MySQL database operations; gained exposure to cloud-based development and cross-platform debugging.

Data Analytics Intern at Rela Multi-specialty Hospital

Designed and developed an interactive dashboard leveraging hospital call center data to provide actionable business insights, including KPIs such as call volume, average handling time, department demand, doctor availability, and appointment slot wait times. Enabled real-time monitoring and data-driven decisions; integrated Tableau visuals and used R for statistics.