Preet Khaturia Email & Phone Number

Houston, Texas, United States

In my current role I have been involved in multiple projects in the O&G and mining sector, enabling client with astute data driven solutions to help them operate efficiently.Ongoing Project: This project is for a client leading in the Oil and Gas sector in Europe and Asia-Pacific. It is related to creating a SaaS product related to oil well planning. Project responsibilities:- Conducting exploratory data analysis to identify trends and patterns, enabling data-driven decision-making for product development.- Working with SMEs to identify the mapping between available data sources and ontological relationships, entities and data models that need to be populated.- Produce Knowledge Graph Database to generate well insights and hazard/risk management.- Led a team of four junior data engineers - which involved coaching and mentorship on data mining techniques.

Hyderabad, Telangana, India

Ahmedabad Area, India

- Developed a Plugin in QGIS for boundary delineation in Agricultural fields through satellite imagery. - Worked on road damage detection (potholes, cracks) for Highway surveillance using image segmentation (FCNs) and detection techniques. - Segmented roads from Indian Driving Dataset (IDD) using Fully convolutional Networks. [https://github.com/preetkhaturia/Road_Segmentation] - Participated in MoveHack 2018, organised by NITI Aayog and provided solution based on the state of the art deep learning algorithm namely, Mask RCNN to detect anomalies present on Indian roads. - Optimized bus routes for Navi Mumbai Municipal Transport (NMMT) system using Travelling salesman algorithm.

Gandhinagar

Thesis: - Detected craters based on contour mapping and shadow/bright region algorithm on Digital Elevation Models (DEMs) and classified craters based on morphology characteristics of craters and cross-sectional profiles.- Proposed a method to automatically detect impact craters on the surface of the Moon by using object detection framework inspired by R-CNN (region proposal) network. - Generated region proposals from ortho images, DEM and slope images using a graph-based segmentation method to limit the search space for a candidate region. - Features are extracted corresponding to each region proposal using 2-D Convolutional Neural Networks (CNN), following VGGNet architecture. - Proposed a post-processing algorithm to eliminate overlapped boxes detected as objects and to optimize the boxes based on its probability score.-Worked as a Teaching assistant in the Electrical Systems lab (ES 431) for Image Processing experiments(Aug 2016- Dec 2016).-Worked as a Teaching assistant for Measure Theory (MA 505) (Jan 2016- April 2016).-Worked as a Teaching assistant for Digital Image Processing (EE 605) and course project mentor for the graduate and undergraduate students (Aug 2017- Dec 2017).-Worked as a Teaching Assistant for Introduction to Analog and Digital Electronics(ES 104).

Pune Area, India

Master Of Technology, Electrical And Electronics Engineering

Bachelor'S Degree, Electrical, Electronic And Communications Engineering Technology/Technician

Nanodegree, Deep Learning