Talentsky, Inc.

August 2020 - present

Data Science Engineer

● Working on skills recommendation engine.

Edgetensor

May 2019 - December 2019

Machine Learning (Computer Vision) Intern

Bloomington, IN, US

● Developed a real-time eye-tracking app using NumPy, OpenCV, CMake, sklearn achieving 20% less pixel error than existing open source projects. ﹤br﹥ ● Automated suspicious activity detection in airport and parking lot videos using TensorFlow, object detection (YOLOv3) and tracking (DeepSORT) cutting monitoring efforts by 80% and expenditure by 50%. ● Bench-marked and evaluated proprietary AI models using pandas, matplotlib identifying cases requiring retraining resulting in cloud computing cost reduction by 70%. ● Engineered an end-to-end smart parking app using NVIDIA Jetson TX2, Docker, NVIDIA DeepStream, Kafka. ● Built and trained deep learning models using Tensorflow to predict eye gaze from image and video datasets to identify distracted drivers and reduce car accidents. ● Communicated results to team members and business clients on a biweekly basis. ● Toolkit: Python, C++, OpenCV, Cmake, Numpy, Scikit-learn, Scipy, Dlib, tKinter, YOLOv3, DeepSORT, NVIDIA Jetson TX2, Pandas, Matplotlib, ctypes, Javascript, Linux, Bash, Git, Docker

Developed a video surveillance solution
September 2019 – October 2019
 

Developed a video surveillance solution to detect stationary people and vehicles in parking lot and airport runway video feed helping cut monitoring costs by 50%.

Developed eye tracking software
May 2019 – August 2019
 

Developed a real-time eye-tracking app using NumPy, OpenCV, CMake, sklearn achieving 20% less pixel error than existing open source projects.

Quadratyx

January 2018 - June 2018

Data Scientist Intern

Hyderabad, Telangana, India

● Deployed a backend Document Analysis service using OpenCV, Tesseract and Django to automate filling of data into MS Excel from scanned bank documents ​reducing man-hours by​ ​50% ​for clients. ● Analyzed and predicted the impact of a trade union’s promotional events on it’s members’ churn rate using ensemble learning ​achieving 83% recall​, ​12% improvement​ over their baseline model. ● Developed load forecasting models for electricity board using GRU and Time Series resulting in ​30% less energy wastage. ● Automated document clustering for banking clients using Tesseract, NLTK and LDA cutting man-hours in half. ● Performed ETL in SQL and set up data pipelines. Created interactive dashboards using Tableau and R Shiny. ●Toolkit: Python, R, SQL, OpenCV, caret, dplyr, Hadoop, Hive, Linux, Statsmodels, Tensorflow, Django, Docker

Indiana University - Bloomington

August 2018 - May 2020

M.S. in Data Science

Bloomington, IN, US

Coursework: Machine Learning, Analysis and Design of Algorithms, Statistics, Deep Learning Systems, Computer Vision, Data Visualization, Elements of A.I., Management and Access of Big Data, Health Informatics

R.V. College of Engineering

August 2013 - June 2017

B.E. in Computer Science

Bengaluru, Karnataka, India

Coursework: Data Structures, Algorithms/DS, Programming in C, Database Management Systems, Object Oriented Programming in C++, Computer Organization and Architecture, Operating Systems, Software Engineering