avatar image

Hasnain Roopawalla

Software Engineering & Machine Learning.

Software Engineer @ Microsoft.


About Me

Summary

Hello! My name is Hasnain and I'm a Software Engineer @ Microsoft in Prague, Czech Republic. I graduated with a Master's Degree in Computer Science from Uppsala University, Sweden in June 2022. My Master Thesis, which I completed @ Scania, focused on increasing the temporal resolution of geospatial data using deep neural networks.

In addition to my love for technology and software, my passions include football, video games, and playing the guitar.

I'm currently working on ShowML - a Python package of Machine Learning Algorithms implemented from scratch, as well as another project, Python TimeIt - a Visual Studio Code extension to view the execution time of any Python function on the fly.

The following sections highlight my skills, experience, and projects.

Highlights

  • Python
  • Dart, Flutter
  • JavaScript
  • MySQL
  • Firebase/Firestore
  • HTML, CSS
  • Android (Java)
  • Flask, FastAPI
  • TensorFlow
  • Keras
  • PyTorch
  • OpenCV
  • MATLAB
  • Cypher, Neo4J

  • Experience

     
     
     
     
     
    Software Engineer @ Microsoft
      Prague, Czech Republic
      Nov. 2022 - Present
    Logo
     
     
     
     
     
    Master Thesis @ Scania
      Södertälje, Sweden
      Jan 2022 - June 2022
    • The thesis focuses on increasing the temporal resolution of trip data for vehicles using a Recurrent Neural Network (RNN)-based architecture, that treats the geospatial data as a time series.
    • Publication and GitHub Repository.
    Logo
    Python
    MySQL
    GAN
    RNN
    PyTorch
    TensorFlow
    LaTeX
     
     
     
     
     
    Software Engineer Intern @ Five AI
      Bristol, United Kingdom
      June 2021 - August 2021
    • Developed a PostFlight Check tool (web app) to systematically examine recorded data from the cars based on user-defined diagnostics (plugins).
    • These diagnostics allow users to identify trends and issues in the raw, sensor data using intuitive plots and dynamic insights.
    • Also worked with:
      • CI/CD (GitLab).
      • Deployment: Terraform, Kubernetes and Docker.
    Logo
    FastAPI
    Python
    HTML
    JavaScript
    AJAX
     
     
     
     
     
    Software Developer @ Hubert
      Stockholm, Sweden
      Nov. 2020 - May 2021
    • Developed a parsing tool for extracting information from candidates' CVs to improve Hubert's virtual interview experience.
    • Designed an algorithm ranging from layout analysis of the document to semantic structure analysis of text objects and chunks.
    • Trained an NER model to parse the candidate's history, qualifications, and skills from the extracted sections (education, experience etc.).
    Logo
    Python
    spaCy
    scikit-learn
    NLTK
    NER
     
     
     
    Machine Learning Intern @ CureAssist
       India (Remote)
      June 2020 - Aug 2020
    • Developed a module for CureAssist's Mobile App to remotely diagnose a patient in under 30 seconds.
    • Built a Graph Database (Neo4J) on the MeSH Dataset to establish semantic relationships among Diseases and Symptoms.
    • Trained a Neural Network on this Graph DB to accurately predict a disease based on the patient's symptom(s).
    Logo
    Neo4J
    Cypher
    Python
    MySQL

    Technical Projects

    ShowML
    ShowML

    Show the Machine Learning Code.

    • A Python package of Machine Learning Algorithms implemented from scratch.
    • The aim of this package is to present the working behind fundamental Machine Learning algorithms in a transparent and modular way.

    Click to read more

    Python TimeIt
    Visual Studio Code Extension - Python TimeIt

    A Visual Studio Code extension to view the execution time of any Python function on the fly.

    • Simply select/highlight the entire function (including import statements if required) with the cursor and trigger the Time It command.
    • The generated function call with the specified arguments and the execution time is inserted as a comment above the selected code.

    Click to read more

    ScribbleNet
    ScribbleNet

    A tool which classifies hand-drawn doodles into distinct classes using a CNN (Convolutional Neural Network).

    Click for a demonstration new! and to read more

    SIH
    Winners @ Smart India Hackathon 2019

    Autonomous Invoice Processing using RPA (Robotic process automation).

    Click to read more

    WeGyan
    Weज्ञान (WeGyan)

    A software which creates a virtual representation of an Institution.

    Click to read more

    Image Captioning
    Image Captioning using Deep Learning

    Generating a caption which describes the contents/scene of an image using LSTMs (Long short-term memory networks).

    Click to read more

    MedBase
    MedBase

    A Web / Android Application for Doctors to store patient history and prescription details along with a fees collection register.

    Click to read more


    Contact

    Drop me an email to get in touch!

    Say Hello