Saurabh Shetty | Data Science - ML

Hi, I'm Saurabh.

My job consists of using my skills as a Data Scientist to optimize and build products at CASHe. I am skilled in all paradigms of the machine learning cycle, such as web scraping, data wrangling, model selection and training, transfer learning, microservices, and cloud development, to build a production-grade application.

After having worked at the research and development wing of one of the world's largest IT firms Infosys Limited as a Specialist Programmer (AI/ML), I'm now working at CASHe as a Senior Data Scientist.

LinkedIn Github

BLOG

A few stories more or less linked with our world, always related to data

Cervical Spine Fracture Detection with FineSpine App

December 2022 - 10 minutes read

Medium post on my project FineSpine App. The competition task is to identify fractures in CT scans of the cervical spine (neck) at both the level of a single vertebrae (location of the fracture/ Fracture Identification) and the entire patient (detection of a fracture/ Fracture Detection). - Read more

Design for Navigation of Underwater Remotely Operated Vehicle

September 2019 - 10 minutes read

This paper presents a design of the mechanical and electronic system for navigating an Underwater ROV providing 6 degrees of Freedom. - Read more

See All

Portfolio

A glimpse of the projects I've been working on

FineSpine App

An end-to-end product built to identify Cervical Spine Fracture in patients

Text Emotion classifier

A Text classification-based methodology to predict the emotion of the given text with model explainability.

Restaurant Recommendation System

A collaborative filter based recommendation system trained by Stochastic Gradient Descent on Yelp business dataset.

Classifying buildings Post Hurricane using Satellite Imagery

Identify flooded regions from satellite imagery.

Santander Customer Satisfaction

Utilizing hundreds of anonymized features to predict if a customer is satisfied or dissatisfied with their banking experience.

Question Answering Deep Learning Model

Finetuned a Google T5 large model on the Question Answering task using the SquadV2 dataset

Sentiment Analysis with Explainable AI

BERT model finetuned for text classification task on five levels of sentiment using Stanford Sentiment tree Bank dataset, achieving a 65% validation accuracy and 92% off-by-one accuracy.

Design and Control of Navigation System for Underwater Remotely Operated Vehicle (ROV)

Remotely operated vehicles (ROV) are remote control underwater robots driven by an individual on the surface. This robot consists of embedded systems, mechanical systems and it is tethered by a series of load carrying umbilical cables which include cables for power, data transmission as well as control signals for the ROV. The purpose of this project is to develop and build the ROV for a specific application and its performance in a simulated environment.

Chem-E-car

Collaborated with the American Institute of Chemical Engineers (AIChE) MIT Pune Student Chapter to build a prototype vehicle using chemical energy.

Camera based error detection system

Engineered an automation system using image detection to identify errors in the data matrix (QR code) and refabricate the error.

Contact

Feel free to contact me for any question. Github. If you want to follow my work, reach me on LinkedIn. Otherwise, send me an email at saurabh.shetty@gmail.com.

Github LinkedIn Mail