John Muchovej, Artificial Intelligence (AI) Developer in Orlando, FL, United States
John Muchovej

Artificial Intelligence (AI) Developer in Orlando, FL, United States

Member since January 3, 2019
John is a recent computer and cognitive science graduate. He primarily works in deep learning, previously focusing on CV and NLP but now focuses on RL and uses ML to interrogate the brain. He's currently a researcher at MIT and Harvard and has been working as an AI engineer for two years—both on self-started projects as well as freelance. He also spends some of his weekends teaching to undergraduate and high school students.
John is now available for hire




Orlando, FL, United States



Preferred Environment

Linux, Emacs, JetBrains, Python, Conda, Docker

The most amazing...

...platform I've developed is a summarization engine for scientific research and visualization tool to enable easier scientific literature navigation.


  • Researcher

    2020 - PRESENT
    Computational Cognitive Development Group, Harvard
    • Developed machine learning models capable of leveraging "common sense reasoning" like human children and adults do.
    Technologies: Python, Deep Learning, Program Induction
  • Researcher

    2019 - PRESENT
    Knowledge Lab, UChicago
    • Analyzed time-series social networks to discern the causal structure, build predictive models, and develop prescriptive models.
    • Developed graphical models that learn over graph-like spaces to derive the causal structure of the graph.
    • Developed graphical models that learn over graph-like spaces and make predictions about the future configurations of the graph.
    • Developed graphical models that learn over graph-like spaces and make prescriptions about how to achieve future configurations of the graph.
    Technologies: Python, Deep Learning, Time-Series Analysis, Graph Learning
  • Researcher

    2019 - PRESENT
    Computational Cognitive Science Group, MIT
    • Reimplemented a "Machine Theory of Mind" (MToM) to be used in future experiments.
    • Analyzed the nuances of said MToM to determine it's learning process and determine if the model learns as portrayed or if the model exploits the latent capabilities of neural networks.
    • Used the MToM in multi-agent settings to model coordination and planning with other agents and humans.
    Technologies: Python, Deep Learning, PyTorch
  • Researcher

    2018 - PRESENT
    NLP Group, UCF
    • Developed a questionnaire for use on Mechanical Turk to quantify how well summaries stick to the content in their source sentences.
    • Developed a training methodology for summarization which generates more faithful summaries to their source articles.
    • Performed analysis on all gathered data to determine how (1)"faithful" (doesn't introduce factual inconsistencies), (2) "covered" (only discusses information in the source sentences), and (3) what kind of sentence fusion generated the most ideal results.
    Technologies: Deep Learning, PyTorch, Pandas, Matplotlib, Python
  • Founder, Past President, Director

    2017 - PRESENT
    • Served as a founding member of the UCF Data Science Team.
    • Designed “Course” curriculum which is offered every semester.
    • Coordinated and maintained [email protected] and Intel sponsorship, $2,500+/semester.
    • Enabled uniform workshops and opened access to non-technical folk by writing an automation framework which runs the majority of managerial tasks for the group.
    • Hosted 20+ workshops, introducing 40-100 students (each) to machine learning and data science techniques.
    • Ran 15+ reading-group meetings where topics on computational cognitive science, implications of current machine learning, and current ML methods were discussed.
    Technologies: Deep Learning, Keras, TensorFlow, PyTorch, Python
  • Researcher

    2018 - 2019
    Center for Research in Computer Vision, UCF
    • Reimplemented multiple, state-of-the-art, action detection and segmentation Neural Networks.
    • Worked towards developing a loss which would augment the training of the above networks by using less data.
    Technologies: Deep Learning, TensorFlow, Python
  • Teaching Assistant, Grader

    2017 - 2019
    • Provided mentorship to 300+ students about the listed technologies. Mentorship included career planning, clarifying technical details, among others.
    • Graded various projects which fall into the above technologies.
    Technologies: Deep Learning, Deep Reinforcement Learning, Natural Language Processing, Computer Vision, Self-Driving Cars, Robotics
  • Graduate Teaching Assistant

    2018 - 2018
    University of Central Florida (UCF)
    • Designed homework successfully reinforcing core deep learning concepts.
    • Coordinated and designed infrastructure for students to complete assignments and final project, using the university’s supercomputer (SLURM batch manager).
    Technologies: Computer Vision, Deep Learning, Keras, TensorFlow, Numpy


  • Forage (Development)

    [Current WIP]
    Discovering and parsing scientific research is challenging for incoming students and those generally ignorant of a given field. Forage is a summarization engine which also models the relationship between fields of research.

  • StarCraft II DeepRL (Development)

    Led a team of five working on replicating and advancing upon the results from DeepMind in the StarCraft II video game. We trained neural networks using deep reinforcement learning (by policy gradient optimization) to achieve similar results to DeepMind across all "mini-games."


  • Languages

    Python 3, HTML, CSS, Sass, Python, Hugo, Julia, R, JavaScript, Go, C++
  • Frameworks

    Bootstrap 4, Hadoop, Spark
  • Libraries/APIs

    PyTorch, Beautiful Soup, Pandas, TensorFlow, Keras, Dask, YouTube API, API
  • Tools

    Jupyter, IPython, Jekyll, Google Cloud AI, Google Cloud Console, AutoML
  • Paradigms

    Data Science, Automation, REST, Continuous Development (CD), Continuous Integration (CI), Continuous Delivery (CD), DevOps, Agile, Object-oriented Programming (OOP), Parallel & Distributed Computing, Object-oriented Design (OOD)
  • Platforms

    Docker, Anaconda, Linux, Arch Linux, Ubuntu, Manjaro Linux, Linux Mint, Amazon Web Services (AWS), CUDA
  • Other

    Deep Learning, Machine Learning, Web Scraping, Artificial Intelligence (AI), Artificial Neural Networks (ANN), Convolutional Neural Networks, Recurrent Neural Networks, Long Short-term Memory, Gated Recurrent Unit (GRU), Neural Networks, Deep Neural Networks, Deep Reinforcement Learning, Ubuntu Server, Front-End Developer, Computer Vision, Data Analysis, Data Analytics, Agile Data Science, Caching, Google Cloud ML, Generative Adversarial Networks (GANs), Google Cloud Natural Language, Natural Language Processing (NLP), GPU Computing, Graphics Processing Unit (GPU)
  • Storage

    Neo4j, PostgreSQL, MySQL, Google Cloud


  • Bachelor of Science degree in Computer Science and Cognitive Science
    2014 - 2019
    University of Central Florida - Florida


  • Deep Reinforcement Learning Nanodegree
  • Natural Language Processing Nanodegree
    JUNE 2018 - PRESENT
  • Computer Vision Nanodegree
    MAY 2018 - PRESENT
  • Self-Driving Car Nanodegree
    MARCH 2018 - PRESENT
  • Robotics Nanodegree
  • Artificial Intellingence Nanodegree
    JULY 2017 - PRESENT
  • Deep Learning Nanodegree
    JUNE 2017 - PRESENT
  • Game Theory
    MARCH 2017 - PRESENT
  • Machine Learning Nanodegree

To view more profiles

Join Toptal
Share it with others