Data Science, AI and new technologies excite me! I love building tools that helps people make data-informed decisions.

Skills

Data Science

  • Keras
  • Scikit-learn
  • PyTorch
  • Tensorflow
  • APIs
  • Bokeh / Seaborn / Plotly

Big Data

  • Apache Spark
  • Hadoop
  • MongoDB

General Development

  • Python
  • Git
  • AWS
  • Django
  • Jenkins
  • Kibana
  • Shell / Bash
  • JSON
  • HTML / CSS

Work Experience (6)

 Munich
Hawk:AI is a real-time transaction monitoring system to indentify and investigate Fraud and Money-Laundering in banking data.
  • Generated artificial banking datasets and trained machine learning models using XGBoost to demonstrate our proof of concept to potential clients. This proved crucial in acquiring first customers (banks).
  • Over a year experience of building and deploying models into the Hawk:AI Java based platform.
  • Made the models explainable using SHAP values, so its predictions could be interpreted by any bank operator.
  • Trained an LSTM-VAE (Variational Autoencoder) using account balance data (time series) and detected outlier accounts. A quick demo and source code can be found <a href=https://github.com/abhmalik/timeseries-clustering-vae>on this link</a>.
  • Extracted 300k time-series features using TA-Lib, trained a VAE and identified suspicious transactions.
  • Manually engineered 80 behavioral features – which expresses changes in usual account activity.
  • Trained a LightGBM model on this data to explain predictions of a Deep Learning model (Autoencoder) which generates such features automatically. These models were used to identify suspicious cases.
  • Built a tool that produces textual explanation for every prediction & helps investigate identified suspicious cases.
  • A demo of the tool can be found <a href=https://github.com/abhmalik/Czech_bank_AML_analysis> on this link</a>.
Developed a tool that automated the whole process of inserting new content, saving 10s of thousands Euros per year.
  • The program systematically edits the HTML code on their Moodle website which was only done manually until now.
  • The tool is written in Python. It uses APIs & MySQL to access database and uses Django for frontend.
Planet Detection Using Machine Learning
  • Developed a novel machine learning method that outperformed current conventional planet detection methods.
  • Method used TSFresh for feature generation and XGBoost. Built a vetting-tool which let people explore predicted planet candidates from Kepler data, a demo can be found on <a href=https://github.com/abhmalik/Exoplanet-Vetting-Tool> this link</a>.
  • Presented a poster at <a href=https://drive.google.com/file/d/1UNUunsj6zufhgomt1TO23Qsgwdu0Q9uz/view?usp=sharing>711th WE-Heraeus-Seminar</a>. A publication on the work is currently in progress.
Executed field evaluation assignments while working in time critical conditions, supervising a 3-Tier team
Preliminary Computational Fluid Dynamics Analysis Of ‘sustor2’
  • Wrote python scripts which enabled a 90% increase in computational efficiency of the program by eliminating GUI
Secured 99.6 and 99.7 percentile in IIT-JEE and AIEEE, two of the biggest national exam in India.
  • Conducted simulations using High Performance Computing (HPC) facility & studied the phenomenon of drag crisis as a part bachelor project.

Volunteer

- Jan 2017
IIT Kanpur
Youngest Invited speaker at IIT Kanpur: Delivered a seminar on Academic Planning for junior students

Education (2)

2020 - 2018
Masters
 Master of Science
Ludwig Maximilian University Munich
2017 - 2012
Bachelors
 Physics & Aerospace Engineering
Indian Institute of Technology (IIT) Kanpur

Awards

2019
Winner
HackaTUM - Official Hackathon of Technical University of Munich
Received first prize for “Smart & Green AI Test System Challenge” by Rohde & Schwarz GmbH & Co KG.
2020
Winner
Tech Xpreience 2020 - Brainport Eindhoven
Won the challenge by Thermo Fisher Scientific Inc. for providing a solution to identify electron microscopy samples using machine learning which explains the model predictions with underlying physical phenomenon.