I am a machine learning engineer with industrial and research exposure on topics such as Computer Vision, Natural Language Processing, Industrial Automation, Autonomous Systems and Robotics. Enthusiastic about breaking into SLAM tech.
Working on specialized LLM (large language modelling) use-cases in the academic editing industry.
Working on various aspects of both Computer Vision and NLP using Machine Learning and Deep Learning with respect to e-commerce stores. My job requires me to build cutting edge AI and Computer Vision Algorithms to improve the visual quality of product listings, along with using NLP to flag suspicious / fraudulent stores.
Working as a Brakes and Air Supply Engineer for the Chennai Metro Phase II tender and handling all the subsystem level requirements and engineering activities for the REM Metro in Montreal, Canada. Have worked on closing special "software-based" solutions in REM, like the unique "Snow-Brake" designed to run at -40° C in Canada.
Worked on developing a Machine Learning Algorithm, under Dr.PS Rana ,for predicting the concrete compressive strength of concrete by training gradient boosting models on the extracted data from concrete samples.
Worked as a part-time robotics engineer/developer, on an Intertial-Aided Navigation system for an Autonomous Collaborative Robot. Developed a ROS-Python package for data acquisition from an IMU-Sensor, without interfacing it with a micro-controller.
Worked at the PAR (Program for Autonomous Robotics) at IIT Delhi, under the guidance of Prof. S.K. Saha. Worked on the electronics and ROS software development of a semi-autonomous robot RoboMuse3XT - for surveillance purposes, and a fully autonomous collaborative mobile robot -RoboMuse V.
Acted as the chief designing officer for an event management startup in Patiala, by managing all graphic design requirements for 30+ events done across Patiala, Jalandhar, New Delhi and Ludhiana. Mentored junior graphic designers to curate creative ideas for the same.
Measuring the concrete's compressive strength is an ever-important task in the civil engineering industry. A collaborative effort has been put up between three institutes to predict the compressive strength of concrete samples using advanced gradient boosting methods, after extracting features such as Water %, Superplasticizer %, etc. A MAPE of 1.2% was overall found after the training of the model.
Developed a mechatronic system for segregating fruits based on their grade by using Spectroscopy and Machine Learning. The fruits are fed to the system and the machine classifies the fruits using Azure Custom Vision API and then using Spectroscopy and a custom gradient boosted ensemble, it grades the fruits into separate classes. The entire system works with the help of an Arduino Mega controller and a Raspberry Pi.
Developed the Software and the Mechatronic system for RoboMuse3XT for helping the robot carry out remote surveillance by controlling it with a PS3 Joystick using ROS. Along with that, I collaborated with another intern to develop a unique robot odometry calibration protocol using a 2D Laser Scanner and the Iterative Close Point method, for RoboMuse V, to calibrate the robot wheel odometry, using a unique path (protocol) for the robot to run.
Developed a Computer Vision solution for counting pushup repetitions while working out from home. The solution is based on the classification of frames using a MobileNetV2 CNN model, specifically trained using transfer learning, with an accuracy of 90% on the test dataset.
Developed a unique ROS Node for acquiring data from the BNO080 IMU (Sparkfun VR IMU) by tapping in the UART-RVC protocol, without using an additional micro-controller. Developed a plug-and-play system using a USB-TTL Converter, that would lead to a saving of Rs. 2000+ per robot manufactured.
Developed an Arduino 2-wheeled Buggy and used CMU's - PocketSphinx and a ROS node to control it using voice commands. Also developed a web app using ROS-Bridge to control the buggy over the internet.
CGPA : 8.62/10
Percentage: 92% (PCM Subjects)
Apart from being a mechatronics engineer, I enjoy working on projects involving graphics, music, movie-making, and creative designing. Other than that, I enjoy traveling to places and documenting them for myself. A music enthusiast, I love talking about and listening to Rock, Hip Hop, and Indian Genres of Music.
Sometimes, when I am not watching TV Documentaries, I try to pick up my guitar and learn something or the other on it. Also catch me on ClubHouse, trying to give talks about career, social media trends, and randomly blabbering about not investing in the DogeCoin.