I'm

Rumman Ahmad

Datascientist , Developer, Machine Learning Engineer
Resume

Personal Introduction

Hi, I'm Rumman Ahmad. Watch this video to know more about me, my journey, and my aspirations.

2

Years

of working experience as a Machine Learning Engineer & a Developer


I'm Rumman Ahmad, a dedicated B.Tech student enrolled in Computer Engineering at the prestigious Jamia Millia Islamia, and will graduate in May 2025.

  • As a MITACS GRI'24 Research Intern, I developed an e-commerce recommendation system using RNN, LSTM, and GRU.
  • Conducted research on PCOS detection using Skip-Connected 2D-CNN and LSTM-1D-CNN, published in MDPI, and submitted to IEEE Conferences.
  • As an ML Engineer at Goalwit Technologies, I developed AI models and deployed them on Azure.


    • Apart from academics and tech, I love football for the excitement it brings. I enjoy restful sleep and exploring the world's beauty.

    Skills

    Web Development

    HTML
    95%
    CSS
    85%
    Javascript
    50%
    React
    70%
    Wordpress
    60%


    Data Science

    Machine Learning
    90%
    Deep Learning
    80%
    Computer Vision
    60%
    Natural language Processing
    50%

    Experience

    Goalwit Technologies

    Machine Learning Engineer             June 2022 - August 2023
    • Experienced Machine Learning Engineer skilled in designing autonomous AI systems for automating predictive models.
    • Developed a precise model utilizing UnderSampling Technique and ML algorithms to predict potential Premium Plan buyers
    • Created a Python Ranking System to efficiently target potential profiles, considering diverse features. Model successfully deployed on Azure.

    Jamia Research Intern

    AI/ML Research Intern             Oct 2022 - March 2023
      Accurate Follicle Detection in PCOS USG Imagery using Skip-Connected 2D-CNN
    • Designed Skip Connections in the 2D-CNN model to improve training efficiency and accuracy.
      Performance Evaluation and Comparison
    • Developed a predictive model that leveraged the combined power of LSTM and 1D-CNN architectures to achieve optimal accuracy. Conducted in-depth studies on metabolic disorder PCOS

    Algoma University (MITACS GR1’24)

    Research Intern             June 2024 - August 2024
    • Project Title: Deep Learning Techniques for Mining Customers’ Sequential Purchase Patterns for E-Commerce Product Recommendation.
    • Gained expertise in Recurrent Neural Networks (RNNs), applying advanced models like LSTM and GRU, achieving a significant improvement in predictive accuracy for sequential data analysis.
    • Optimized Deep Learning techniques to analyze historical customer data, identifying patterns that improved e-commerce product recommendations.
    • Analyzed performance using the Amazon All Beauty dataset containing over 5K+ user ratings to evaluate the effectiveness of the system.
    • Evaluate the performance of the recommendation system in terms of accuracy, efficiency, and scalability

    My Projects

    • Transformers
    • Major Projects
    • GAN Projects
    • CNN Architecture
    • ML Projects
    • Recommendation System
    Developed a Deep Convolutional Generative Adversarial Network (DCGAN) using PyTorch.
    Link
    • Implemented a custom Generator and Discriminator neural network architecture.
    • Trained the GAN on the MNIST dataset to generate realistic handwritten digit images.
    • Monitored and visualized the training process using TensorBoard.
    Wasserstein GAN with Gradient Penalty (WGAN-GP) Implementation
    Link
    • Created a unique Generator and Discriminator neural network design from scratch.
    • Incorporated the Wasserstein distance as the loss function to improve training stability.
    • Successfully trained the WGAN-GP model for 5 epochs, achieving significant improvements in the generated image quality over time.
    Implemented a Deep learning model using InceptionV3 architecture to classify
    Link
    • Developed a deep learning model for dog breed classification by leveraging the InceptionV3, architecture.
    • Implemented a feature extraction function that utilized the InceptionV3 model
    • Utilized GlobalAveragePooling for feature representation.
    Facial Recognition-Based Music Recommendation System
    Link
    • Developed a facial expression recognition system using AI (Neural Networks) that recommends songs based on detected emotions.
    • Upon detecting facial expressions, the system directs users to Spotify to access recommended songs.
    Deep Learning Model for Handwritten Digit Recognition using InceptionNet
    Link
    • Implemented the InceptionNet architecture, incorporating custom InceptionBlocks to extract hierarchical features from the input images effectively.
    • Generated a confusion matrix to analyze the model's performance on individual classes and identify potential areas for improvement.
    Advancing MNIST Digit Recognition with ResNet-based CNNs
    Link
    • Utilized the BasicBlock structure to build a robust and efficient CNN model for digit recognition
    • Utilized the CrossEntropyLoss function as the loss criterion and the Stochastic Gradient Descent (SGD) optimizer with momentum for efficient model optimization
    VGG16-based Convolutional Neural Networks
    Link
    • Preprocessed the CIFAR-100 dataset with image resizing and normalization.
    • Achieved a high accuracy rate on the validation set, demonstrating the model's effectiveness.
    MNIST Digit Recognition using EfficientNet
    Link
    • Implemented an EfficientNet-based deep learning model for digit recognition.
    • Achieved high accuracy on the test set, demonstrating the model's effectiveness for digit recognition. Visualized the confusion matrix to assess model performance.
    WINEQUALITY PREDICTION
    Link
    • Predicted the quality of wine using DecisionTreeClassifier
    • Used PCA for selecting the features,Clustering is done with the help of KMeans,DBSCAN and pruning the dataset while selecting least ccp_alpha parameter.
    PREDICTING DIABETES
    Link
    • Made a model using ML algorithm which has been trained and tested using model_selection module
    • Converted the dataset into standard value using standardization method . Predicted the tested value which gives an accuracy of 80%
    Personalized Recommendation System for E-Commerce
    Link
    • Developed a personalized recommendation system using graph-based filtering and biased random walk for user interactions.
    • Utilized Word2Vec embeddings and evaluated the model with precision, recall, novelty, diversity, and CTR.
    • The system aimed to deliver relevant, diverse, and novel product suggestions for e-commerce.
    Movie Recommendation System using Collaborative Filtering
    Link
    • It predicts user preferences based on the behavior of similar users.
    • Data visualization techniques were employed using Matplotlib and Seaborn to gain insights into the data
    • A Pivot Table was utilized to recommend movies based on user preferences and improve the model.
    MobileSR with Transformers
    Link
    • Designed and implemented a hybrid Mobile Super-Resolution (MobileSR) model integrating convolutional layers and transformers (GPT-2) for enhanced image reconstruction.
    • Optimized the training process using Cosine Annealing Learning Rate Scheduler and Adam optimizer for efficient convergence.
    • Employed L1 loss and evaluation metrics such as SSIM and PSNR for high-fidelity image restoration.

    Roles and Achievements

    • Google Developer Student Club - JMI

      Oct. 2022 – May 2024

      • ML / AI Team Lead: Directed a three-day ML/AI workshop with 100+ student attendees.
      • Tech Fest: Hosted a Kaggle competition with participation from over 50 teams.

    • MITACS Global Link Research Internship 2024 Scholar

      Acknowledged for excellence in research.

    • Fatima Predoctoral Fellowship 2024

      Awarded the prestigious fellowship for academic and research achievements.

    • CodeChef Starters 25 Division 3 (Rated)

      Global Rank: 759

    • HackInit Hackathon

      3rd Position: Secured 3rd position in the hackathon organized by JMI.

    Publications

    • R. Ahmad, L. A. Maghrabi, Ishfaq Ahmad Khaja, L. A. Maghrabi, and M. Ahmad

      “SMOTE-Based Automated PCOS Prediction Using Lightweight Deep Learning Models”, Diagnostics, vol. 14, no. 19, pp. 2225–2225, Oct. 2024.

    • R. Ahmad, P. Pvan, P. Jadhav, A. Sharma

      “Exploring the Integration of Wearable Sensor Technologies in Professional Sports for Enhanced Athletic Performance and Injury Prevention” (under review in IEEE Xplore).

    • D. Dhablia, Ranjith Singh, K. KaheR, Ashish Verma, R. Ahmad

      “The Evolution of Personalized Medicine in the Age of Big Data and Advanced Analytics” (under review in IEEE Xplore).

    Letter of Recommendations

    • Dr. Mohammad Amjad, Head and Professor, Department of Computer Engineering, Jamia Millia Islamia View
    • Dr. Musheer Ahmad, Assistant Professor, Department of Computer Engineering, Jamia Millia Islamia View
    • Mr. Hammad Rahman, Founder and CEO, Goalwit Technologies Pvt Ltd View

    CONTACT ME

    Country:

    India

    Call me:

    +91 7982364612

    Mail me:

    rummanahmad05@gamil.com

    Follow me:

    © Rumman Ahmad, All Right Reserved.