Home About Skills Work Ex Projects Achievements Extra-Curricular

Who I am

Hi there! I am Harshita Gupta
I graduated from BITS Pilani. I am a data enthusiast and currently working as a Graduate Trainee Engineer in FosFor(data product) at Larsen & Turbo Infotech, Bengaluru. I believe in the art of story telling from the data which can solve the major part of the problem.
I love reading, learning, and researching about how AI has revolutionized the human-like tasks, making it possible to learn from its mistakes.

profile_photo

Education

BITS Pilani, Pilani Campus

2018 - 2022

I was enrolled in a integrated four year programme, pursuing B.E. in Civil Engineering.

CGPA : 7.8/10

RLB Memorial School, Lucknow

2015 - 2017 (Class XII)

School was affilated to Central Board of Secondary Education (CBSE).

Marks : 96.6%

SMC School, Sultanpur

2015 (Class X)

School was affilated to Central Board of Secondary Education (CBSE).

CGPA : 9.6/10


Skills

Python 80%
SQL 75%
C++ 45%
Java 20%
HTML/CSS 15%
JS 5%
Pandas 75%
Numpy 70%
Scikit-Learn 60%
Keras 40%
Streamlit 35%
Pytorch 25%

Courses

  • Machine Learning*
  • Basics of Data Warehousing*
  • DBMS*
  • Data Mining`
  • Neural Networks and Fuzzy Logic`
  • Applied Statistical Methods`
  • Information Retrival`
  • Object Oriented Programming`
  • Probability ans Statistics`
  • Linear Algebra`

    • `Academic Courses

    *Online Courses

    Work Experience

    Graduate Trainee Engineer

    Larsen & Turbo Infotech | July 2022 – Present

    Data Science Intern

    Subex - AI Labs | Jan 2022 – June 2022

    Supervisor: Asif Salim, Research Scientist

    Simbox Fraud Management System

  • Analyzed and featured engineered the CDR using aggregate functions for a better interpretation of the data
  • Built custom graph data from tabular data and GCN using Pytorch Geometric for label cleaning to provide a relevance score for each record
  • Built a logistic classifier over GCN to predict if the call is legitimate which provided an accuracy of 89%

    • CounterFactual Explanations (Explainable AI)

  • Introduced CFE metrics: diversity, sparsity, and proximity to evaluate the objective function in PSO
  • Added information value parameter for each feature to improvise CFE with sparsity control

    • Python Developer

    William ONeil, Pvt. Ltd. | Jul 2021 – Dec 2021

    Supervisor: Manjunatha Gowda, Lead QA Engineer

  • Added news features in the MSI product, a sections to show the latest news, NSE-BSE bulk/block, NSE insiders data, and derived fiscal year formula
  • Developed pipelines to scrape data using feedparser, selenium beautiful soup, cleaned and feature engineered the data, built loggers, and integration with SQL servers
  • Visualized the code for better understanding of the data and optimization of the existing codes so that it could be easily integrated with the cloud servers in python

    • Computer Vision Research Intern

    CSIR-CEERI, Pilani | Feb 2021 – Apr 2021

    Mentor : Dr. Dhiraj Sangwan, Principal Scientist

  • Survey of publicly available datasets for Human Pose Estimation
  • Explored various models for Human Pose Detection and identifying human joint key points
  • Developed a feedback mechanism using Vector Cosine Angle 2D for the analysis of correct yoga posture to model the Virtual Yoga Trainer

    • Summer Intern

    Ultra Tech Cement - IT | May 2020 – Jul 2020

  • Developed an application that keeps track of the several online payments made by the user online
  • Built a sign-in page that allows user to enter their detail to make an account using the basics of HTML for structuring and CSS for styling the page

    • Projects

    Achievements

    5 Terre

    Secured rank 12 in
    Amazon ML Challenge
    (July-2021), hosted by HackerEarth.

    Norther Lights

    Top 27% in
    Customer Attrition Prediction,
    Summer AI Challenge
    (May-2021) organized by Kharagpur Data Analytics Group at Kaggle.

    Norther Lights

    Secured rank 3 in
    Pneumonia Classification Challenge
    by Segmind (Apr-2021) hosted by Dphi.

    Norther Lights

    Secured rank 4 in
    Attentive AI Internship Hiring Challenge (Jan-21) based on an image classification model for satellite hosted by Dockship.

    Norther Lights

    Rank 12 in Burn Out prediction Rate of Employees (Dec-2020) hosted by HackerEarth.

    Norther Lights

    Top 20 in AitaHack 1.0 a Data Science Hackathon (Dec-2020) by Aitahub.

    Norther Lights

    Secured rank 2 in Weekend Hackathon #15: Melanoma Tumor Size Prediction (Aug-2020) hosted by MachineHack.

    Extra-Curricular

    Community Contributor,
    Dphi-Democratizing Data Science Learning

    (POR)

    Jul 2021 – Sept-2021

    Assisted 200+ budding data science enthusiasts by solving their queries, projects, competitions on Data Science related questions in the Deep Learning Bootcamp arranged by Dphi.

    Event Coordinator,
    Hindi Activity Society, Oasis, BITS Pilani

    (POR)

    Sep 2019 – Nov 2019

    Increased the footfall of ~300 participation from various colleges in Filmy Muqabala is a fun-based quizzing event with questions geared towards Bollywood movies.

    Pondering over the methods to clean data in ML-problems.

    (Publication)

    19 Feb, 2021

    This article covers major steps of visualizing, analyzing, pre-processing/transforming raw data, validating, and hyperparameter tuning to make predictions for improving model performance in machine learning problems. (Dockship)

    Computer Vision DevCon 2020,
    Association of Data Scientist

    (Conference)

    13 - 14 Aug, 2020

    Selected for a two-day virtual Tech Talk for AI influencers among 500+ participants from across the world along with some best-selected paper presentations.