A Straight, To-the-Point Portfolio

Your Photo

Aagam Bakliwal

Master's at Cornell University

Experience

Aug 2024 - Jan 2025
Alpha Capital Logo

Data Engineering Intern

Alpha Capital

  • Leading the development and execution of core trading strategies using Spark for high-volume data processing and Amazon EMRFS for efficient data storage, driving growth for the firm’s trading operations
  • Building statistical models to analyze volatility patterns in the options market, utilizing Kafka for continuous data ingestion and streaming, leading to optimized strategies that increased profitability and reduced risk exposure
Mar 2024 - Jun 2024
Beryl Drugs Logo

Software Developer Intern

Beryl Drugs

  • Implemented a predictive inventory system for Small Volume Injectables using ARIMA and GRUs for forecasting, deployed via Kubernetes, and visualized in Tableau, achieving a 3.5% reduction in excess inventory waste
  • Deployed a ML pipeline using Flask for backend processing and Docker for containerization, applying SVMs and K-means clustering to analyze consumption patterns and optimize resource allocation, leading to a projected 1.7% revenue boost
May 2023 - Jul 2023
Mastercard Logo

Software Developer Intern

Mastercard

  • Implemented CI/CD pipelines, handled ETL processes for large datasets and engineered predictive models using SVR and XGBoost for Mastercard’s Carbon Calculator, achieving an adjusted R2 of 0.91
  • Ideated an ML-driven solution for enhancing security in real-time payments for Mastercard’s nationwide innovation challenge, leading the team to secure 2nd place, with the project now under patent review
Dec 2022 - Sep 2023
Department of Defense Logo

Software Intern

Department of Defense

  • Developed software utilizing GRUs and RNNs for predicting key characteristics of solid rocket propellants such as Burn Rate, Pressure and Thrust, achieving a MSE of 0.034
  • Engineered GAN and RL-based simulations to uncover new solid rocket propellant formulations, successfully identifying 2 mixtures moving forward to empirical testing
  • Designed and deployed computer vision models to accurately identify and create bounding boxes on metal casting defects achieving 95.4% precision, resulting in over 30% reduction in time to completion
Nov 2021 - Dec 2022
TalentParse Logo

Co-founder - Acquired Startup

TalentParse

  • Developed a framework using Node.js, Express.js, and MySQL, containerized with Docker, and implemented a novel ML ensemble with Association Mining and Knowledge Graphs to resolve orphan entities in resumes, achieving 86% accuracy
  • Led the negotiations that resulted in the sale of our novel algorithm, culminating in its acquisition by the AI-based recruitment platform Rezooomex

Projects

Intelligent Melanoma Detection Automation System

(AWS EC2, Flask, RESTful API) | Jan 24 - May 24

  • Implemented a Flask-based RESTful API, deployed on AWS EC2 instance, and integrated with an Amazon RDS SQL database for secure patient data storage and retrieval
  • Developed a scalable, lightweight ML solution for melanoma detection using an ensemble of custom CNNs, CBAMs, and Vision Transformers, achieving 91.2% accuracy and winning the best paper award at the IC2SDT'24 conference

Distributed Learning System for Healthcare Imaging

(Kubernetes, React, AWS, PyTorch) | Dec 23 - Apr 24

Third Best Software Project in the University

  • Developed a highly accurate computer vision model using an ensemble of CNNs, achieving a mean average precision (mAP) of 0.367 on the VinBigData chest X-ray dataset, a 3% improvement over the best published model
  • Developed a React-based dashboard for healthcare providers to track model performance and data flow, and implemented the federated learning infrastructure on AWS using Kubernetes for scalable deployment

American Depository Receipt Correlation

(Tensorflow, Flask, React) | Jan 23 - May 23

  • Developed an algorithmic trading solution that identified the correlation between American Depository Receipts (ADRs) and their Indian stock counterparts, which was later acquired by Alpha Capital
  • Developed a React-based web interface integrated with Flask that implemented a machine learning ensemble using LSTMs, GRUs, and regression algorithms to process daily stock data, achieving a 72% predictive accuracy for stock price correlations

Skills

Programming Languages
Python C Dart Swift HTML CSS JavaScript
Databases & Processing
MySQL SQLite PostgreSQL MongoDB Spark Hadoop AWS Sagemaker Athena
Data Visualization
Plotly Matplotlib Seaborn Tableau Power BI
Packages and Tools
Scikit-learn NumPy Pandas TensorFlow PyTorch Git Docker Kubernetes CI/CD (Jenkins)
Web Development
Node.js Express.js React Flask REST