About Us

Who Am I?

Hi I'm Ashish Gupta I am an Enterprenure and have expertise in AI Software developement. I love to find solutions using AI. I have worked at WadhwaniAI where I was a part of Maternal Child and Newborn(MNCH), team.Previously I have worked with IIT KGP as research assitant. I have also worked at Signzy as AI SDE-2 and previously worked with Advance computing lab at IIT Madras.

Other then Academics,I have also worked Computer Society of India ,BIT Student Chapter Head. I have worked as a operation coordinator at Quizophilic India and as a sponsorship coordinator at Cavorts Sports Fest BIT Mesra. And Sposorship and managenet coordinator at Surbhi Music Academy.

I am an AI enthusiast and also in search between connection of AI with Humans.Having Innovation and curious mind, I am constantly on the lookout for solutions to existent problems in our everyday life and look forward to creating something that helps to improve the lives of billions of people across the world. I always see desiging AI as teaching a small child to perform some task. I am always curious to learn different dimensions of AI like backpropogation,Dataset Preparation,Generalization(Bias,etc.) and model designing.

In my spare time, I love making music(lost when guitar or piano in hand and singing loud).I love reading books and listining podcast and playing games. I love to read about the patterns exist in this world. I believe in the concept of lifelong learning.I can occasionally be found blogging on Medium.

Experience

Work Experience

Associate Machine Learning Scientist,WadhwaniAIAug 22 -Current

Associate Machine Learning Scientist WadhwaniAI

  • Holds ownership of Pregnancy report Stratification (PRS) Project.
  • Being touchpoint between Client and different teams for technical design and end to end deployment.
  • Achieved specificity and sensitivity of 80% and 70% respectively in short interval of time with ready to deploy model.
  • Explored different LLM's on Indian Languages for PRS usecases.
  • Developed AI-based newborn virtual weighing machine with other anthropometric parameters.
  • To run-on low-end devices such as mobiles, reduced AI model size by 92% and improve processing time by 91%, improving efficiency and prediction speed.
  • Increased prediction accuracy by 20% by combining visual and tabular data.
  • Conducted various model analyses (attention, weight distribution).
  • Showcased analytical skills and ability to interpret complex AI model behavior.
  • Build and Manage data collection pipeline with automatic data quality checks, that reduced data collection wastage by 30%.
  • Performed POC’s for MNCH projects such as Growth Faltering, PRS.

Core AI SDE-2, Signzy Jan20 - Jan22

  • Manages ownership of diverse microservices and AI products, deploying on AWS, Azure, and on-prem end-to-end solutions.
  • Awarded best team Player, as Worked closely with multiple teams and roles to make sure the deliveries happen & are executed.
  • Being a touchpoint for clients and Multiple Vendors.
  • Specialized in advanced technologies: Deep Face Authentication (1:N match), face match, and 1:1 face detection, OCR.
  • Implemented solutions for digital tampering challenges, including screenshot and copy-paste manipulation, and copy-move detection.
  • Applied secure masking techniques for Aadhaar and EkYC forms, utilizing OCR and segmentation.
  • Led the design and implementation of extraction solutions with Named Entity Recognition (NER).
  • Reduced weight/usage on advance API's by developing image quality analysis, which rejects 30% of bad quality images.
  • Research Fellow, IIT Madras 2019

    I am developing 3D reconstruction techniques using Machine learning.It's a part of my ongoing work on 3D geometry .We are using ShapeNet and Pascal VOC dataset for the same.

    Junior ML Engineer Omdena 2019

    • A team of 40 technocrate from 23 countries is working for social welfare to find out the real time solution for forest fire outbreak in Brazilian Forest using AI. The team will work for 8 weeks (2 months) i.e Oct 2019 to Dec 2019 to find out the solution . Our work includes Semantic Segmentation Problems with image classification.
    • Project Link
    • Blog Link

    Research Fellow Under Professor Arun K. Pujari 2019

    Under my research under Professor Arun K. Pujari Vice Chancellor,Central University Of Rajasthan,Ajmer I worked on 3D object reconstruction. Its like give me a image of object and I will construct the 3D of that object. We have used Shapenet 3D and Pascal Voc dataset with Deep Residual Networks(CNN).

    Backend/Desktop Application Developement Intern at Orbet Softech May 2018-July-2018

    At Orbet I worked on developing Student Online Test System with Marks Evalution Using Fuzzy Logic(Online Portal) with JSP,MySQL,HTML,CSS and develop various small Desktop Application in Java.

    Marketing Intern,Content Creator,Designer at Unfold Craft May 2016- June 2018

    At Unfold Craft I worked on Content Creator and also as a artist in Video Based on Friendship.As marketing intern I have worked in finding Sponsors for Independence Day video and also given music in a video on Raksha Bandhan

    Head/Convenor Computer Society Of India BIT Mesra Chapter 2017-Currently Active

    Under CSI I have organised various seminers.talks,activities,classes.I have also started AI club in the college.In CSI I have a greate team and we work dedicately and some of our activities are Codeing Competations like Freshers Code and Seminars by techniest on embedding systems and classes include python programming and Under CSI we have also stated a mentorship programme which helps student to develop project on various topics under seniors.

    Operations Coordinator Under Quizophilic India 2017-2018

    At Quiophilic my work is to handle management and analyze and improve organizational processes, and work to improve quality, productivity and efficiency for different Qizzing Events.I am also worked as in the main team for Enigma Quiz and IPL Auction I have also worked as Sponsorship Coordintor. I also worked as head in organising techQuiz in BIT Mesra.

    Education

    Education

    I am persuing Bechlor of Engineering in Computer Science Engineering From Birla Institute Of Technology.

    I have done my 12th in Rajasthan Board from St. Jaorge Academy.

    My Projects aka Work

    Recent Projects

    3D Object Reconstruction

    Deep Learning & Computer Vision(3D)

    Object Pose Estimation(6 DOF)

    Rotation and translation Matrices using CNN

    NIM Game Using AI Hurastics

    Java Desktop Application using Swings

    Botanical Flower Detector with Web Search Facility

    Deep Learning(CNN) with OpenCV and Google API for web Search

    Sentiment Analysis

    Deep Learning(RNN) with imdb dataset

    Online Test System with Marks Evalution Using Fuzzy Logic(Online Portal)

    Java Server Pages with MySql for backend and HTML,CSS Frontand
    Read

    Recent Blog

    HTML5 Bootstrap Template by colorlib.com
    May 30, 2019

    Constructing Visual Hull for 3D object reconstruction

    Visual Hulls plays a crucial role in 3D reconstruction and 3D reconstruction now a days are in hot zone.Lets learn and delve deep into the algorithms.

    HTML5 Bootstrap Template by colorlib.com
    June 15,2019 | AI Journey

    A Year of Learning AI


    I have began AI a year before and during this journey I find various difficulties.I want to share all the resources that I find useful.

    Publications

    Publications

    A Novel Multi-Task Learning Approach for Context-Sensitive Compound Type Identification in Sanskrit

    Ashish Gupta, Jivnesh Sandhan ,Hrishikesh Terdalkar, Tushar Sandhan, Suvendu Samanta, Laxmidhar Behera, Pawan Goyal

    COLING 2022



    NurtureNet: A Multi-task Video-based Approach for Newborn Anthropometry

    CVPR Workshop-CVPM 2024

    Paper Yet to present

    Project page