• Led a multidisciplinary team to develop specialized 3D medical visualization software for personalized pelvic implantplanning, utilizing Scientific Python and Qt, in accordance with IEC-62304.• Managed the entire project lifecycle (SLDC), including design, implementation, testing, and production.• Proposed and developed proof-of-concept AI technologies to integrate into company workflow.• Developed a nnUNet-based deep learning module written in PyTorch for automated pelvic CT imagesegmentation, reducing implant design time by 80%.• Successfully containerized the deep learning based module using docker, facilitating the deployment.• Training a full stack deep learning model to automatically segment pelvic CT scans with metal artifactsusing the MONAI framework.• Implemented advanced 3D mathematics algorithm to improve usability of the application.• Extended existing C++ and OpenGL code in a legacy software to incorporate the AI segmentationresult.• Implemented various computer graphics algorithms such as Marching Cubes and Ramer-Douglas-Peucker (RDP).• Explored various mesh decimation algorithms in Python to optimize the size of exported STL files.• Brainstormed and designed the UI of the software with various stakeholders using Figma to ensureuser-friendliness.• Designed the user interface of the application using the Qt framework.• Provided comprehensive documentation and user training to facilitate ease of use for end-users.• Extensively used various design patterns such as observer design pattern in python.• Ensured adherence to the SOLID principles throughout the entire project.• Gained experience working with various medical imaging formats such as DICOM, NIfTI, and NRRD.Skills: Software Design · Software Project Management · Transformers. nnUNet · Python. PyTorch. IEC-62304. Github. Poetry. Software Development Life Cycle (SLDC). Agile. OpenGL. C++. Qt. Docker. DICOM. MONAI.

1. Collaborated in developing individual features and integrated them to the software. Gained experience in object oriented programming and dynamic substructuring.2. Explored different options for speeding up matrix operation.

Experimental implementation, modeling and control of a lower arm prosthesis for the Cybathlon challenge.Implemented a machine learning algorithm to recognise various grasping intents of the amputee using EMG signals. The algorithm was successfully deployed on the hardware.Designed an estimation algorithm, based on Extended Kalman filter, to track the joint parameters for better controller design.Exploring the dynamics of motor torque variations, based on the unique contact surface properties of the grasped object, was a fascinating journey. This intricate relationship was meticulously simulated and visualized using Simulink, revealing nuanced insights into the system's behavior.Tools: MATLAB. Simulink

Investigated the effect of intermetallics on the strength of the ball bond connection in a power electronic chip when subjected to a thermal gradient. Examined the variation of the interfacial energy during different reliabilty test performed on a power electronic chip, in order to observe the critical regions.Explored different mathematical methods for modeling adhesion in power electronic chips.Skills: Extended Finite Element Method (XFEM), Molecular DynamicsTools: Python, Linux

Reliability analysis of different ''physics of failure'' phenomenon of power electronic packages, using advanced techniques such as Monte Carlo and Markov modeling.Conducted detailed data analysis and statistical evaluations to assess the relationship between key variables in various simulation scenarios.Skills: Finite Element Method(FEM), Reliability AnalysisTools: Python, Linux

• Led the development and execution of computational fluid dynamics (CFD) simulation of the pressuredeveloped in a low voltage switchboard in case a short circuit happens. Also suggested design updatesto mitigate the unwanted catastrophic effects.• Simulated the fracture dynamics of a carbon reinforced plastic made bus-bar support in ANSYS, inorder to obtain the amount of force that the busbar can safely support without fracturing.• Analysed the safe amount of mechanical force that can be exerted on different components of a lowvoltage switchboard by simulating the effect using ANSYS.• Provided insight in how varying the material type and dimension of a busbar can effect the forcebetween two parallel busbar by conducting a electromagnetic simulation in ANSYS.• Worked closely with cross-functional teams to validate and enhance models, providing crucial simulation data for timely design reviews of power switchboards with different ratings.• Leveraged extensive Python scripting within ANSYS to automate complex workflow processes, significantly reducing manual effort and improving overall efficiency.• Brainstormed and implemented innovative design of various components of the low voltage switchboard using CAD software such as PRO-E and SOLIDWORKS.• Led the development and execution of test such as short-circuit test, IP-65, lifting test, thermal testetc on the switchboards of various rating in accordance with IEC-61439 and successfully got the designcertified for a multi-million dollar project for the GCC market.• Tech Stack: Python, ANSYS-Workbench, ANSYS-Fluent, PRO-E, Solidworks