Shivansh Verma

Experience

• 

  Jio Platforms Limited (JPL)

  May 2019 - Jul 2019

  Machine Learning Engineer

  • Developed Deep Learning models that identified CAPTCHAs and broke them using Transfer learning and leveraged image processing frameworks such as OpenCV. • Leveraged pre-trained neural networks from TensorFlow and YOLOv3 for several object detection projects.• Leveraged YOLO (Darkflow implementation) for object detection.• Developed an Object Detection model that identified people and their dress colours to help with faster Person of Interest Identification.• Analyzed and cleaned several large datasets to be used for training purposes for deep learning projects.• Helped deploy several machine learning models into cloud frameworks like Docker and Kubernetes. Show less

• 

  Phamoss Pvt Ltd

  Sept 2020 - Jan 2021

  Software Development Engineer

  • Developed the entirety of the backend with a small team of 3 for a nascent startup.• Integrated modern NodeJs frameworks such as ExpressJs and Sequelize to production level code and ensure optimum performance.• Worked on MERN Stack ( MySQL, ExpressJs, ReactJs, NodeJs ).• Developed core backend functionalities for applications on REST standard.• Deployed the entire application on AWS Infrastructure and ensured a HA and Resilient application using ASGs and ALBs.• Integrated CD/CI capabilities in AWS for faster development times. Show less

• 

  Deloitte

  Jan 2021 - May 2021

  Technology Intern

  • Worked on REST frameworks in Spring Java and Python• Trained on Okta and AWS • Created several REST APIs and micro services for various Databases.

• 

  Deloitte

  Jul 2021 - Jul 2022

  Advisory System Analyst

  • Designed and integrated Single-Sign-On capabilities for clients having 100+ active applications.• Helped clients understand modern identity and access management practices and implement them.• Extensively worked on Okta to provide SSO capabilities and modernize applications.• Initially started off as an intern campus hire and was offered a fulltime role after completion of the internship.

• 

  Northeastern University

  Nov 2022 - now

  Research Assistant

  • Research assistant in a large scale data mining project involving scraping the Securities and Exchange Commission Filings and doing text analysis for risk factors for 25,000+ companies having 1,000,000+ documents.• Systematized 500,000+ articles and extracted complex parts of the subjects for further Natural Language Processing.• Increased the accuracy of item identification from 70% to 93% by using Fuzzy Logic.• Improvised methods to create automatic prompts from documents for Large Language Models.• Processed 1 TB of employment data from 1.1 billion profiles in 4.5 million companies worldwide to understand trends. • Leveraged advanced analytics techniques to derive actionable insights to understand workforce and employment dynamics.• Engineered effective data solutions with high throughput to perform quantitative analysis on large datasets.• Designed and executed parallel computing techniques in Python, resulting in a remarkable 30% reduction in computation time. Show less

• 

  WHOOP

  Jul 2023 - Dec 2023

  Machine Learning Engineer

  • Applied findings from research publications to develop Neural Networks that predicted various health metrics using Siamese Networks.• Built novel multi modal Neural Networks on PyTorch that effectively utilized Nvidia GPUs both on EC2 and AWS SageMaker.• Performed transfer learning on ResNet18 by leveraging AWS SageMaker for large scale training on 1TB of health data.• Decreased the absolute error of prediction by 45% by incorporating different modalities along with using ResNet18 and AlexNet.• Engineered robust system designs for training on large datasets by leveraging multi-GPU and parallel training on SageMaker. Show less