profile-pic
Vetted Talent

Charu Jain

Vetted Talent

Computer vision engineer with 12 years of experience in the IT industry.

1) Languages: Python, Java, C, Scala

2) Machine learning: Statistical Modelling, Bayesian statistics, ML algorithms, Neural networks(Library - Tensorflow 2), LLM, Langchain, RAG

3)Web development-Angular JS, Web services (REST, SOAP)

For details about my projects please refer to the section below.

  • Role

    Senior Module Leader

  • Years of Experience

    11 years

Skillsets

  • object detection
  • rag
  • LLM
  • LangChain
  • Docker
  • Computer Vision
  • Statistical shape modelling
  • TensorFlow
  • Springboot
  • REST
  • OpenMP
  • OpenCV
  • Open3D
  • Deep Learning
  • CUDA
  • Classification
  • Bayesian Statistics
  • AngularJS
  • Scala
  • C++
  • Segmentation
  • Kubernetes
  • Keras
  • Python
  • Java
  • SOAP

Vetted For

10Skills
  • Roles & Skills
  • Results
  • Details
  • icon-skill_image
    Front End Developer (3D Image Processing)AI Screening
  • 46%
    icon-arrow-down
  • Skills assessed :cornerstone.js, DICOM, itk.js, NIFTI, FrontEnd, vtk.js, ES6, JavaScript, Python, React Js
  • Score: 46/100

Professional Summary

11Years
  • Jan, 2024 - Jul, 20251 yr 6 months

    Senior Module Leader

    Motherson Technologies And Services
  • Dec, 2021 - Oct, 20231 yr 10 months

    Senior Computer Vision Engineer

    Vehant Technologies
  • Jan, 2018 - Dec, 2018 11 months

    API Developer

    Barclays
  • Jul, 2011 - Jan, 20175 yr 6 months

    Systems Engineer

    Infosys
  • Jan, 2017 - Dec, 2017 11 months

    Senior Module Lead

    Sopra Steria

Applications & Tools Known

  • icon-tool

    Tensorflow

  • icon-tool

    RAML

  • icon-tool

    REST API

  • icon-tool

    SciPy

  • icon-tool

    Spring Boot

  • icon-tool

    VTK

  • icon-tool

    3dslicer

  • icon-tool

    Scalismo

  • icon-tool

    Computer Vision

  • icon-tool

    NumPy

  • icon-tool

    Java

  • icon-tool

    Python

  • icon-tool

    Keras

  • icon-tool

    Google Kubernetes Engine

  • icon-tool

    BigPanda

  • icon-tool

    C++

  • icon-tool

    Scala

  • icon-tool

    Angular

  • icon-tool

    CUDA

  • icon-tool

    Azure Machine Learning Studio

Work History

11Years

Senior Module Leader

Motherson Technologies And Services
Jan, 2024 - Jul, 20251 yr 6 months
    Designed solutions and led a team to implement multiple projects including Document Image Comparison (automated visual comparison of engineering drawings using computer vision), Bill of Material Extraction (LLM and RAG for contextual understanding), Crimp Cross-section Analysis (image registration and segmentation using UNet and VAE), and Forklift Safety (YOLO-based tracking system deployed on Jetson Nano). Applications deployed on AWS Sagemaker and tracked with MLFlow.

Senior Computer Vision Engineer

Vehant Technologies
Dec, 2021 - Oct, 20231 yr 10 months
    Conceptualized and prototyped product enhancements for Under Vehicle Scanning System (UVSS), including Bayesian coherent point drift algorithm for point cloud registration, statistical shape modeling of video data, FLANN and KDTree for nearest neighbour search, deep learning for image classification and segmentation (ResNet, UNet), and performance optimization using CUDA, OpenMP, LAPACK, and MAGMA.

API Developer

Barclays
Jan, 2018 - Dec, 2018 11 months
    Developed new microservice for Pingit mobile app using REST framework and Spring Boot. Managed deployment and microservices using Kubernetes.

Senior Module Lead

Sopra Steria
Jan, 2017 - Dec, 2017 11 months
    Developed web application (Online Quotation Tool) using Javascript and J2EE technologies following scrum methodology.

Systems Engineer

Infosys
Jul, 2011 - Jan, 20175 yr 6 months
    Worked on internet banking modernization, new business strategy management tool, and internet banking regression for clients National Australian Bank and American Express.

Major Projects

4Projects

Image Segmentation of Tooth in CBCT Image Using Iterative 3D U-Net Training to Build a Tooth SSM (Master Thesis)

    Automated 3D tooth segmentation from CBCT images using 3D U-Net for semantic segmentation to build statistical shape models. Evaluated model quality based on generalisation, specificity, and compactness.

3D Slicer Extension

    Developed a user-friendly UI extension using the open source 3D Slicer framework for editing image labels in 3D.

Combining Statistical Shape Model of Complete Tooth from CBCT Images with Laser Scan Data of Its Crown (Master Project)

    Created a hybrid 3D model by combining full morphology from CBCT images and details from laser scans using Metropolis Hastings algorithm with iterative closest point proposals.

QA Bot to Read Document (Coursera)

    Built a generative AI application using LangChain and RAG for document question answering.

Education

  • B.Tech, Information Technology

    Jaypee University Of Information Technology (2011)
  • M.Sc, Computer Science

    University Of Basel (2021)

Certifications

  • MuleSoft Certi Developer - API Design Associate (RAML 1.0) (cid:221) Oct 2018 - Oct 2020

AI-interview Questions & Answers

I have I have a total nine years of experience. My first six years of experience is in software engineering background where I have worked on front on web development and deployment activities. After that, I did my master's for two and a half years and I specialized in machine learning. And one and a half years of my duration in master's was working in the project for medical on medical images, which included DICOM files and creation of tools related to image segmentation. And yes. And after that, I have worked in behind technologies where I have worked on point cloud image processing. Yes. And that's it.

Well, I have built an app for a web app that allows to edit 3D segmentations using 3D slicer software. So, 3D slicer has an API using which I was able to build web components which were able to access the machine learning algorithm and it was able to predict the segmentation on the web app and we were also able to edit using the APIs, we were able to apply web components which were able to edit the images on the screen. Using 3D slicer API, I was able to do it.

In what ways would ES6 and ReactJS aid you in building web apps? Well, ReactJS will provide me web components which would help me in building the web application. ES6 provides a standardized way of doing the same. So using the web components provided by ReactJS, I will be able to provide interactivity with the application.

Cornerstone.js is basically a JavaScript library which helps in interacting, which helps in working with medical images. So this kind of a framework would, and it helps in visualizing the 3D images in different directions in different forms. So such a JavaScript library in combination with React.js can be used for loading DICOM files or NIFTY files and visualizing them in 3D space in XYZ directions, in XY plane along YZ plane along ZY plane, or as a 3D rendering format.

The project structures. Well, the project structure that I use is that I keep the constant variables in a separate file. And I organize my data, my project in such a way that it is that the components are reusable and they are upgradable. And I and all the components are also unit tested as well as system tested end to end so as to build a scalable platform for a front end application.

Can you provide an example of complex front-end feature you developed in technology stack that you used? Well, this was a long time back when I was working in web application development for a banking application. We were able to build a login feature where there's no complex front-end feature that I've developed but just normal web components for login and for transitioning for two different application where you roll down and you're able to dynamically load more content on the screen.

Consider the below Python function that loads DICOM image and visualizes 3D segmentation using what potential issues can you spot and how can you suggest to improve? So it is able to read the file and return the pixel array. Well, there should be a check whether the data which is loaded is present or not. So there needs to be an exception which needs to be added. So that there is a fallback in case the file is not present. There needs to be an exception rule in load image while it's loading the file. That's all.

Review the following charges against a student that uses consumption as a premedical measure and whether you might be wrong with it. Load image should be present before enabling the element Thank you for watching!

In the current project in Vihant where I was working, we had to alter the image which was displayed in 2D to 3D and also it had to highlight the 3D points which were present on the screen where a foreign object was present, so working with 3D data using a new library was a difficult task so the existing 2D image loading had to be changed to 3D image loading as well as rendering that on a 3D view box, that was a challenging situation

What are the strategies you have employed to ensure a small functional GPU dependent browser? Well, GPU dependent browser application, I have used for 3D rendering of images or 3D reconstructed images that I was working with. So, we used libraries which also provided functions for working on GPU. So, the idea was that if GPU is available, then it works on GPU, else there is a fallback on CPU as well, in case GPU capability is not present. To ensure smooth functioning of GPU dependent browser application, we had to make it independent from the hardware that was being used. So, the idea was that we were using open source libraries so that they are not dependent on vendor specific hardware configuration. So, we were using open source libraries for GPU based applications.