• M.Sc. in Electrical Engineering at University of Stuttgart (2020-2023)
• Master’s Thesis: “Point Cloud Up-Sampling and Domain Adaptation” at Valeo Detection Systems
• Backend Developer at Oracle Inc, India (2015-2019)
• BE in Electrical and Electronics Engineering, VTU, India (2011-2015) (Recommendation 1) (Recommendation 2) (Appreciation)
• Bachelor's Thesis: "Multicolor RGB LED Display Board"

• Spearheaded pioneering research initiatives in-cabin driver and passenger monitoring systems using RGB and Depth sensors.
• Implemented state-of-the-art models such as Yolov8 and v7, surpassing performance benchmarks on open-source COCO datasets.
• Orchestrated dataset recording plans, ensuring comprehensive coverage of boundary conditions and enhancing model robustness.
• Utilized advanced tools including Kubernetes, ClearML, MLOps, Docker, GIT, Confluence, and Jira to streamline project workflows and collaboration.
• Skills: Python, SOTA Research, Kubernetes, Docker, MLOPs, ViT.

• Engineered an advanced deep learning model for LiDAR point cloud up-sampling, achieving an eightfold increase in point generation and significantly improving resolution.
• Optimized data management workflows on Google Cloud Platform (GCP), resulting in a 35% reduction in processing time and enhanced operational efficiency.
• Generated simulated LiDAR point cloud traces using carmaker software, facilitating comprehensive testing and validation processes.
• Skills: LiDAR, Python, Pytorch, TensorFlow, Keras, ViT, GANs, CNN, Open3D, Point Cloud Processing, 3D Reconstruction, CARLA, ROS, State Estimation, CarMaker, GCP Cloud.

• Contributed to the Engineering System/SW Comfort Functions team within the Honda Lane project for L1 and L2 autonomous driver monitoring.
• Led the enhancement efforts for the WATSH plugin, which included implementing various features such as lane keeping, lane departure warning, and emergency lane keeping based on road edge detection and oncoming vehicle detection.
• Played a pivotal role in optimizing the plugin's functionality, resulting in reduction in false positive alerts and enhancing overall driver safety features.
• Skills: Git, CI/CD, Image Processing, C++.

• Contributed to the build, test, and deployment of a web-based enterprise solution impacting over 100,000 users in 66 countries.
• Utilized tools such as Oracle Business Intelligence (OBI), SQL, Power BI, and Oracle Data Integrator(ODI) for data collection and processing.
• Led test automation efforts for the Fusion T&L software release using Selenium and effective data collection and processing.
• Achieved a 30% reduction in testing time for Fusion T&L software releases.
• Demonstrated ability to collaborate within cross-functional teams and deliver impactful solutions.
• Skills: SQL, BI, Oracle Cloud, Testing.

• Developed a conditional GAN-based model for portrait image editing, improving the quality of the edited images by 10% compared to the original paper.
• Implemented style transfer using gram-matrix loss and component transfer to improve the realism in the generated image.

• Trained CNN-based VAE models for generating facial images, achieving a 90% accuracy rate in image recognition tasks.
• Implemented Grad-CAM and Guided Grad-CAM techniques, improving model interpretability and achieving a 15% increase in object tracking accuracy.
• Skilled in applying Grad-CAM and Guided Grad-CAM to VAE and U-net models for tasks such as semantic segmentation, and object recognition.

• Designed and implemented a customer-side Python-based desktop application for processing and displaying the diagnosis of eSIM inside Porsche Taycan Telecommunication Control Unit (TCU).
• Facilitated the enhancement of eSIM communication systems in Porsche Taycan vehicles through comprehensive testing and the development of a user-friendly diagnostic tool, ensuring efficient troubleshooting and maintenance.
• Demonstrated proficiency in software development and quality assurance processes within the automotive industry.
• Skills: Python, ADAS, Tkinter, ADB, Low level Serial Communication Protocol.

• Led a multidisciplinary team to the finals of the competition by implementing Yolov5 on provided datasets, achieving an outstanding test accuracy of 86.96%.
• Developed innovative image segmentation techniques for anomaly detection, showcasing expertise in computer vision and machine learning.
• Skills: Deep Learning, Python Programming, Rapid prototyping, Team Management, Resource allocation, SOTA Research.

The aim of the project was to track the progress of work at a construction site.

• Track the progress of concreting, reinforcement, and form-work at a construction site using video data.
• The code implemented an algorithm to detect and draw bounding boxes over the input video and display the percentage of progress.

The aim of this project was to build a windows application to detect edges from the incoming images from 2 cameras to interpret the depth of the image and measure the CPU cycles and improve it.

• The final project picked input image files and output another image with the edges details.
• The project was successfully completed in time and it was built without using OpenCV or any other Impage Processing libraries and all the functions were written from scratch in C++ and Emphasis was given for Code Clarity, Ease of use & Robustness.

• Implemented algorithms like Q-Learning, Deep Q-Learning, Lambda-Q, Monte-Carlo and Markov models.

• Successfully programmed ESP32 using ESP IDF with python and download a pre-trained model to detect single words using TensorFlow Lite.
• Successfully Programmed ESP32 to work as a Smart socket using MQTT-Mosquito broker running on Amazon Ec2 Instance.
• Programming Languages and tools: Python, TensorFlow Lite, Linux, PCB designing, TCP/IP Protocol, and ESP-IDF

• Led winning teams in consecutive hackathons, showcasing problem-solving prowess and leadership skills.
• Analyzed time series data of a room heater to predict room occupancy, enhancing skills in data analysis and modeling.
• Utilized Infineon depth sensor camera to generate 3D point cloud from raw data, demonstrating proficiency in advanced technologies.
• Secured 1st place in the hackathon by demonstrating excellence in problem-solving, data analysis, and utilization of advanced technologies.

• Converting an old PC into home server to use it as personal cloud data storage device with starting capacity of 9TB.
• OS used is Ubuntu, services include drive, email server, video streaming, user accounts and upload and download of large files.

• Built ESP32 based smart socket using MQTT server running on AWS EC2 instance.
• Tested ESP-CAM and TensorflowLite
• Building Smart LED bulbs and opening electronic equipment like motor, fan, CPU.
• SMT soldering upto SOIP package