Novella Alvina

Novella Alvina

Data Science Enthusiast

👋🏼 Hello there, I’m Novella!

👩🏻‍💻 I’m a graduate student at the University of Utah.

🔬 I’m currently part of LL4MA, instructed by Prof. Tucker Hermans.

🚀 Interest

  • Machine Learning
  • Artificial Intelligence
  • Natural Language Processing
  • Deep Learning
  • Data Mining
  • Database
  • Computer Vision

📚 Education

Master of Science in Computing, Data Management and Analytics Aug 2024 — present

University of Utah Salt Lake City, UT

Bachelor of Science, Data Science Jan 2022 — Dec 2023

University of Utah Salt Lake City, UT

Associate Degree of Science, Computer Science and Mathematics Sep 2019 — Jun 2021

University of Durham Durham, UK

📝 Featured Publications

Yixuan Huang, Tucker Hermans, \underline{Novella Alvina}, Mohanraj Devendran Shanthi. “Fail2Progress: Learning from Failures with Stein Variational Gradient Descent for Robot Manipulation Tasks”. Robotics: Science and Systems (RSS). 2025 [in preparation]

  • Co-author on a research paper investigating learning from failures based on an active-learning approach, posing failure-informed data generation in simulation as a variational inference problem
  • Helped in analysis of experimental findings to generate figures and video demonstration of the planning approach and failure reasoning methodologies to communicate results and theoretical concepts
  • Adapt and evaluate object detection and segmentation model in robotic tasks, contributing to the refinement of failure-handling strategies

💼 Experience

Research Assistant for LL4MA lab Aug 2024 - present

University of Utah

  • Adapt and optimized motion planning, grasping algorithms, and data heuristic collection code for the STRETCH robot using Isaac Sim, ensuring seamless simulation functionality
  • Design a scalable, parallelized framework integrating motion planning and data collection, improving efficiency, streamlining workflows, and advancing collaborative robotics research.
  • Enhanced the localization and mapping capabilities of the STRETCH robot by integrating the SplaTAM SLAM system with advanced data fusion techniques, improving navigation accuracy by 10%, and optimizing 3D Gaussian splat mapping to reduce rendering time by 10% through efficient processing of RGB-D data from the RealSense camera

Teaching Assistant for Machine Learning course Aug 2023 — Dec 2023

University of Utah Salt Lake City, UT

  • Mentored 60+ students from Bachelor to PhD levels, guiding the implementation of advanced machine learning models for real-world challenges, including TensorFlow, scikit-learn, Keras, PyTorch and XGBoost
  • Assessed assignments and developed a comprehensive curriculum, resulting in a 20% increase in student engagement and achievement

Financial Systems Analyst Intern Jun 2021 — Aug 2021

Harapan Bangsa Foundation Tangerang, Indonesia

  • Implemented automated tracking for 1000+ employee loans using Python, cutting processing time by 20% and errors by 30%
  • Streamlined the salary database, achieving a 30% increase in data accuracy and a 15% reduction in payroll processing time
  • Deployed a monthly expense prediction model with a 5% margin of error, enhancing budget management and providing data presentation for decision-making in the finance department

💻 Project

Deep Learning Based Chatbot, Developer Oct 2024 - Dec 2024

  • Developed a deep learning-powered chatbot using advanced NLP techniques (BERT, Rasa NLU, DiagloGPT) for personalized recipe recommendations, focusing on intent recognition and entity extraction to enhance user interaction
  • Evaluated chatbot performance using metrics such as Intent and Entity Recognition Accuracy, BLEU, and ROUGE for semantic similarity

Tech: Python, BERT, Rasa NLU, DiagloGPT, NLTK

Web App Development, Back-end Developer and Database Engineer Jan 2023 — Jan 2024

  • Enhanced book recommendation and content genre generator precision by 20% through automated machine learning and improved NLP preprocessing, boosting accuracy by 25%, reducing loading times by 30%, and increasing user engagement by 25%.
  • Engineered a robust API using Django’s Rest Framework, optimizing communication between frontend and backend systems; increased system integration efficiency by 30% and reduced latency by 50 milliseconds.

Tech: Python, Django, SQL, Docker, Git, Github, Gitlab, AWS, PostgreSQL, Postman, tensorFlow, nltk

AI Pacman Agent Game, Developer Aug 2023 — Dec 2023

  • Developed and implemented an AI for a Pac-Man game using Python and Tkinter, integrating reinforcement learning (Q-learning) and advanced pathfinding algorithms (A* and BFS) to optimize gameplay, resulting in a 90% win rate over conventional systems.
  • Conducted a comprehensive analysis of AI behavior under various game scenarios to fine-tune parameters and improve the decision matrix using MinMax, ExpectiMax, and Alpha-Beta Pruning methods, leading to a a 30% improvement in ghost evasion tactics and significantly increased game longevity and scoring through detailed behavioral analysis and parameter optimization.

Tech: Python, Tkinter

D3 Data Visualization, Developer Aug 2023 — Jan 2024

  • Optimized Python-based data pipeline with NLP preprocessing, increasing efficiency by 50%. Implemented diverging bar charts for sentiment analysis on artists’ songs over multiple years to perform exploratory data analysis
  • Constructed a spider chart for various song attributes, to analyze of song characteristics leading to 30% increase in decision clarity

Tech: Python, d3, Javascript, html, css, pandas, numpy, sci-kit learn, nltk, Github

Web Server Learning Management System(LMS), Database Engineer Jan 2023 — May 2023

  • Developed a resilient web server database for LMS resulting in a 30% reduction in administrative time. Wrote complex SQL queries and employed LINQ API library in C# for data processing and increased operational efficiency

Tech: SQL, MySQL, C#, .NET Core

Machine Learning Libraries, Developer Aug 2022 — Dec 2022

  • Constructed a comprehensive supervised ML training program from scratch, including decision trees, ensemble learning models, linear regression, artificial neural networks (ANN), support vector machines (SVM), and perceptron
  • Achieved a 30% enhancement in output prediction accuracy across diverse data models and algorithms

Tech: Python, pandas, numpy, matplotlib, seaborn

🛠️ Skill

Courses

✅ Machine Learning

✅ Artificial Intelligence

✅ Deep Learning

✅ Natural Language Processing

✅ Database Systems

✅ Data Mining

✅ Statistical Analysis R Programming

Other Skills

✅ R

✅ Google Colab

✅ Java

✅ C

✅ Microsoft Office Suite