Dennis Loevlie

Dennis Loevlie

Followers of Dennis Loevlie1000 followers
location of Dennis LoevliePittsburgh, Pennsylvania, United States

Connect with Dennis Loevlie to Send Message

Connect

Connect with Dennis Loevlie to Send Message

Connect
  • Timeline

  • About me

    Machine Learning Researcher in the Tufts ML Group

  • Education

    • Carnegie Mellon University

      2019 - 2021
      M.S. - Chemical Engineering 3.91
    • Freeport Area Senior High School

      2011 - 2015

      Activities and Societies: PJAS (Pennsylvania Junior Academy of Science), PRSEF(Pittsburgh Regional Science and Engineering Fair), Mock Trial, Key Club, Varsity Soccer, Varsity Track, Executive Student Council, Chorus, Jazz Band, DECA

    • West Virginia University

      2016 - 2019
      Bachelor of Science - BS Chemical Engineering Cum Laude

      Activities and Societies: Undergraduate Research, AIChE, Honors College During my time at West Virginia University, I became extremely passionate about the fields of chemical and software engineering. I began research with Dr. Fernando Lima’s group during my second semester at WVU and continued this research until August 2019. Additionally, I was the Vice President of West Virginia University’s chapter of AIChE.

    • La Roche University

      2015 - 2016
      Chemistry

      Activities and Societies: Men’s soccer team, CheM scholars, Honors College In the fall of 2015, I started at La Roche University and played NCAA Division III Soccer. During my time at La Roche, I did research with Dr. Boyzm on pigment experimentation where I presented my first poster on the undergraduate level. I was a member of the CheM Scholars program and the Honors College. During my time at La Roche, I realized I wanted to pursue a degree in engineering rather than chemistry which led to my decision to transfer during my sophomore year.

    • Tufts University

      2024 -
      Master of Science - MS Computer Science
  • Experience

    • West Virginia University

      Jan 2019 - Aug 2019

      My experience as an undergraduate researcher was extremely broad. I worked under Prof. Lima in the CODES group. The main projects that I worked on are listed below:• My first project involved the EPA’s chemical process sustainability software, GREENSCOPE, and its potential applications. • My second and more complex project involved mathematical modeling and optimization of large-scale ion transport membranes as an alternative to cryogenic air separation units for oxygen production. This involved MATLAB programming, ASPEN Plus modeling, as well as, building a custom mathematical model for the membrane in the simulation software SimCentral. Show less

      • Undergraduate Researcher

        Apr 2017 - Aug 2019
      • Student Grader

        Jan 2019 - May 2019
    • Carnegie Mellon University

      Dec 2019 - Jan 2021
      Graduate Student Researcher

      During my time at CMU, I was highly active in Dr. John Kitchin's research group. A few of my major contributions are listed below:• Recreated image analysis tools in Python (originally in Mathmatica) to be faster and more intuitive. • Developed a Python package, nb_search, to efficiently sort through, locate and open Jupyter notebook files. • Trained a CNN to accurately predict the relative amount of residuals deposited in the bottom of an experimental well.• Regressed parameters and used them to cluster different bimetallic catalysts (used in the Hydrogen Evolution Reaction) into 5 distinct groupings. Show less

    • AithELITE

      Jan 2021 - Jun 2021
      Data Scientist/Software Engineer

      • Developed web scraping scripts using Beautifulsoup and Selenium to automate data retrieval and keep the data up to date• Automated the data/feature engineering with Numpy and Pandas• Applied Machine Learning algorithms using Numpy and Sklearn to generate intelligent predictions from the data• Built and maintained relational (MySQL) and graph (Neo4J) databases, hosted on AWS• Built the front and backend of the NovAItion EliteAI website with Django, hosted on AWS

    • University of Pittsburgh

      Jun 2021 - Dec 2022
      Research Assistant

      At the University of Pittsburgh, I researched under Prof. Mpourmpakis (Yanni) in the CANELa group. My main contributions are listed below:• Developed a novel method for calculating cohesive energies of metal nanoparticles that led to a 71% improvement to the previous model. • Applied quantum/statistical mechanics through Density Functional Theory, Boltzmann statistics, and mixed integer optimization using the genetic algorithm to better understand structure property relationships for metal nanoparticles. • Organized and presented in several machine learning and software development meetings within the CANELa research group. Notable presentations include: version control using git/GitHub, an introduction to PyTorch and logistic regression, creating interactive visualizations with plotly and dash, and developing custom Python packages. • Mentored members of the group about ML fundamentals, building models with PyTorch, and evaluating ML results. Show less

    • KEF Robotics

      Jan 2023 - Aug 2024

      At KEF Robotics, I spearheaded the development and integration of advanced deep learning solutions, focusing on Universal Image Segmentation, object detection and Monocular Depth Prediction. My responsibilities spanned the entire pipeline: from data preprocessing to the selection, implementation, optimization, and integration of state-of-the-art (SoTA) models using PyTorch and PyTorch Lightning.Project 1: I led the machine learning efforts for a client-focused ground-vehicle/tethered UAS reconnaissance project, showcasing KEF's aerial autonomy within an Unreal Engine simulation. Utilizing the Mask2Former model with a Swin-L backbone, I was able to accurately segment challenging obstacles like power lines. This implementation not only marked a significant stride in infrared image segmentation but also set a new benchmark for KEF's capabilities. A pull request I made on the Hugging Face Transformers library enabling PyTorch 2.0 integration led to a 23% boost in inference speed, and a custom ResNet50 backbone further enhanced speed by 45% with a slight 4% accuracy trade-off.Project 2: I have the privilege of leading the NAMC 10x24 initiative at KEF Robotics, a significant project with a budget of $500,000 and a one-year timeline that combines advanced robotics with artificial intelligence to enhance our mission-critical operations. While my primary technical focus lies in the integration of object detection and state-of-the-art Monocular depth capabilities onto the two separate commercial drones for real-world hazard detection and avoidance, my overarching responsibility is to lead and manage a dedicated team of five engineers. Show less

      • Senior Computer Vision and Machine Learning Engineer

        May 2024 - Aug 2024
      • Computer Vision Engineer

        Jan 2023 - May 2024
    • Tufts University

      Aug 2024 - now
      ML Researcher in the Hughes Lab

      Working on multiple instance learning for 3D neuro-images. This work is directly applicable to video understanding problems. Currently developing ways to improve attention. Generally, my research is applicable to machine learning problems with limited amounts of labeled data.

  • Licenses & Certifications

  • Honors & Awards

    • Awarded to Dennis Loevlie
      Second Place Overall The Pitt Challenge Oct 2022 Awarded second place overall in the annual Pitt Challenge hackathon.
    • Awarded to Dennis Loevlie
      Third Place Poster Award Carnegie Mellon University, Chemical Engineering Masters Student Association (ChEMSA) Nov 2020 Carnegie Mellon University, Research Forum, Poster Competition
    • Awarded to Dennis Loevlie
      Presidential Honors College Graduate West Virginia University Honors College Aug 2019
    • Awarded to Dennis Loevlie
      NSF Summer Research Grant National Science Foundation Jun 2019
    • Awarded to Dennis Loevlie
      Dean’s Honors West Virginia University May 2019
    • Awarded to Dennis Loevlie
      2nd Place Poster Award in Computational Modeling and Controls American Institute of Chemical Engineers Oct 2018
    • Awarded to Dennis Loevlie
      Dean’s Honors West Virginia University May 2018
    • Awarded to Dennis Loevlie
      NSF Summer Research Grant National Science Foundation May 2018
    • Awarded to Dennis Loevlie
      Dean’s Honors West Virginia University Dec 2017
    • Awarded to Dennis Loevlie
      Dean’s Honors West Virginia University May 2017
    • Awarded to Dennis Loevlie
      Dean’s Honors La Roche University May 2016
    • Awarded to Dennis Loevlie
      Dean’s Honors La Roche University Dec 2015