Experience
Epsilon Solutions
Sr.Python Engineer(Part-time) | Feb/22 - Feb/23
- Lead a team by preparing coding & quality check guidelines which resulted in a 10/10 customer satisfaction score.
- Dissected 20+ proprietary protocols used in industrial automation by reverse engineering networking traffic and analyzing the byte code of the embedded devices, while having no prior knowledge of dissection.
- Led project to completion 5 days ahead of schedule while meeting product specifications and reducing the cost of product research and development by 50%.
- Automated development environment deployment, reducing setup time from 2mins to <30secs, allowing faster response to customer requests.
L&T Technology Services
Automation Engineer | Jul/18 - Nov/21
- Awarded ”Star of the Month - May, 2021” for developing the POC for a project that earned a $1 million contract.
- Provided IT/OT architectural support for the 20+ PLC OEMs in the File Parsing project & designed software standards.
- Identified the ”Golden” batch of production from data analyzing approx millions of data from more than 100 batches in an Industrial IOT project, leading to a 20% decrease in production costs.
- Designed production data visualization dashboards (Key-insights, efficiency & uptime metrics) on ThingWorx IOT platform & Django.
- Conceptualized custom widgets for showing performance matrices, including average cycle time and total throughput, by modifying the JavaScript source code of the application.
Academic
Carleton University
M.A.Sc. (System and Computer Engineering) | Jan/22-Current
- Research Topic: Simulteneous Localization and Mapping (SLAM)
- MITACS Internship with Airy3D
- Board of Governor : Sep/22 - Aug/23
L. D. College of Engineering
B.E.(Instrumentation & Control) | 2014 - 2018
- Semi-Finalist in Ni-Yantra by National Instruments
- Contingent Leader in Mood Indigo - IIT Bombay
- Sponsorship head & Core team Member in Sahitya Sarita - 2018
Publications
A Survey on Visual SLAM Algorithms Compatible for 3D Space Reconstruction and Navigation
IEEE International Conference on Consumer Electronics (ICCE) - 2023
3D SLAM survey of conventional and deep learning based methods
Motor Imagery Brain Activity Recognition through Data Augmentation using DC-GANs and Mu-Sigma
IEEE Sensors Conference - 2022
Improving accuracy performance of MI-BCI EEG signals classification dataset by using GAN and Mu-Sigma methods to overcome overfitting.
Projects
These are the projects I've worked upon for different objectives, may be some are from course projects, some are free-lancing projects I did for Ph.D. students from non-coding fields to make their life easier.
Instance Segmentation using Mask R-CNN on Pascal VOC dataset
Preparing instance segmented dataset for training such model is laboursome task. So, this project was a kind of an experiment to check how pre-trained Mask R-CNN on ImageNet dataset artistically can be used to predict object mask on easily labelled bounding box Pascal VOC dataset.
Vehicle ID mapping from multi-drone captured footages
The task was to map vehicles with same ID passing in different frame captured by multiple drones settled at fixed time offset. Input data was Camara footage with vehicle id and a csv file with local frame cordinates of vehicle ID for each footage, now because of the time offset, Vehicle ID n in footage F wouldn't be the same in footage F+1. So, I required to match those vehicles ids using mathematical formulas and regression. Because of the confidentiality, source code of this project is not available.