Ayush Yadav Email & Phone Number

New York, Ny, Us

Bengaluru, Karnataka, India

- C++ Software Engineer in the Scenario Simulation team, responsible for developing the Scenario Simulation Engine (SSE) for large-scale discrete event simulation workflows.- Added support to generate Ready-To-Run C++ package actors from Simulink models. These packages support co-simulation between RoadRunner (https://in.mathworks.com/products/roadrunner.html)and MATLAB without any Simulink overhead (eliminating expensive Simulink runtime and memory overhead). With these Ready-To-Run… Show more - C++ Software Engineer in the Scenario Simulation team, responsible for developing the Scenario Simulation Engine (SSE) for large-scale discrete event simulation workflows.- Added support to generate Ready-To-Run C++ package actors from Simulink models. These packages support co-simulation between RoadRunner (https://in.mathworks.com/products/roadrunner.html)and MATLAB without any Simulink overhead (eliminating expensive Simulink runtime and memory overhead). With these Ready-To-Run actors, we 3x-ed the average simulation pace. - Actively mentoring Engineers from the EDG group on projects related to Scenario Simulation Team- Actively working on simulation projects for the SimEvents product. (https://in.mathworks.com/products/simevents.html) Show less

Bengaluru, Karnataka, India

- Worked on multiple C++ projects with the Scenario Simulation team. Added support for multiple coordinate systems (Projected, Geocentric, Geographic and Tangential Coordinate Reference Systems) in the Scenario Simulation Engine.- Led the entire Engineering Development Group as a Group Lead to provide technical assistance to customers on MathWorks product suites.

Bangalore Urban, Karnataka, India

Designed and implemented C++ APIs for users to upload custom Simulation Scenarios and Simulation Maps from MATLAB to the Scenario Simulation Engine via gRPC.I also provided technical assistance to customers on MathWorks product suites as a part of the Engineer Development Group.

Bengaluru, Karnataka, India

Teaching Assistant for SM 102 - Linear Algebra under Prof. Manisha Kulkarni.

Bengaluru, Karnataka, India

- Worked on a project to detect false positive liver lesion segmentations from CAT scans (.nrrd files) - Generated Radiomics features from CAT scans and developed multiple classical machine learning models to detect false positive segmentations. - Integrated the fp detection models as a post processing step to the larger liver lesion detection and segmentation pipeline.

Bengaluru, Karnataka, India

Teaching assistant for GEN 511 - Machine Learning under Prof. Dinesh Babu Jayagopi.

- I worked on the FORUM 4.0 project which aims to automate the moderation requirements of user comments in online public forums (public forum example: www.ndr.de)- Our work was published at KOVENS 2021: forumBERT: Topic Adaptation and Classification of Contextualized Forum Comments in German. (https://aclanthology.org/2021.konvens-1.17)- To solve the problem of contextualized comment moderation, I proposed and implemented forum-BERT, which uses two weight-shared BERT models and a late… Show more - I worked on the FORUM 4.0 project which aims to automate the moderation requirements of user comments in online public forums (public forum example: www.ndr.de)- Our work was published at KOVENS 2021: forumBERT: Topic Adaptation and Classification of Contextualized Forum Comments in German. (https://aclanthology.org/2021.konvens-1.17)- To solve the problem of contextualized comment moderation, I proposed and implemented forum-BERT, which uses two weight-shared BERT models and a late fusion technique to determine the quality and relevance of a comment on a forum post.- Our solution achieved state-of-the-art classification results on two publicly available datasets (One Million Posts dataset, NDR dataset). Show less

Bengaluru, Karnataka, India

Teaching Assistant for the GEN 512 - Mathematics for Machine Learning course under Prof. V Ramasubramanian.

Karnataka, India

- Given an input depth image of a hand gesture, I finetuned a voxel-to-voxel 3D convolutional neural network to estimate 3D hand keypoints - Developed a pipeline to map these 3D hand keypoints to a virtual avatar in Blender.- Reduced manual effort in gesture creation from 20 minutes to less than 30 seconds by providing accurate hand keypoints.

Master Of Science - Ms, Computer Science

Integrated Mtech, Computer Science, 3.73/4

Activities and Societies: AI/ML specialization. Captain, Institute Basketball Team Core Member, Institute Sports Committee.

12Th, Physics, 96.5%