Christopher Kok Email & Phone Number

- Developed an AI-driven learning and development solution using fine tuned large-language models (LLMs), grounded with L&D expertise through traditional ML techniques, knowledge graphs, and expert-in-the-loop knowledge retrieval.- Architected and deployed scalable machine learning pipelines on cloud platforms (AWS, Azure) using Docker and Kubernetes, achieving a 83% increase in model training efficiency and reducing deployment times by 67%.- Designed an OWASP-compliant LLM anonymization system to prevent sensitive data leaks (97+% of PII masked).- Secured strategic partnerships with MIT, Microsoft, and Accenture; enhancing product reach and credibility. Also established early interest from government and Fortune 500, including a six-figure sales contract in development.

Utilizing machine learning to make teaching scalable and personalized. My research goals include analyzing students performance to determine prior knowledge (strengths and weaknesses), helping instructors understand these analyses, and providing personalized recommendations for learning material or further action.My current research project allows instructors to easily create rubrics for students' exams/quizzes, grade their answers, and provide personalized feedback. This is done through a multi-stage, semi-supervised machine learning process for clustering students' answers. Instructors are better able to articulate good/bad example answers, plus understand the overall class progress and pain points. From this, they can also effectively tweak their future instruction.With the help of Professor Xu Wang, I published 2 papers in the first year to an international edtech conference (Learning@Scale2023, Copenhagen). I also led a team of 20 graders for EECS 376 (700+ students). Established clear rubrics, improving consistency.

Accelerating the adoption of AI/ML at Collins Aerospace (Raytheon Technologies).I'm leading a passionate team of 25 engineers and technical fellows to establish the taxonomy, best practices, technical road map, and suggested tools for AI/ML development. Beyond that, we're creating an internal network of all projects, teams, and individuals to streamline collaboration. We secured $250K in funding for this effort. I'm also co-leading the AI/ML community of practice distributing knowledge and creating connections between our 1000+ members.AI/ML advocacy aside, I'm involved in the end-to-end application development involving several machine learning and data analysis research contracts. I explore and implement machine learning research on various million-dollar aerospace projects (from using natural language processing for software requirement generation to utilizing meta-learning for in-flight activity recognition and decision making). My most recent contract is in partnership with the FAA (Federal Aviation Administration) to develop a highly scalable, cloud-based infrastructure to crowdsource and automate weather observations. On this project, we efficiently process and utilize about 360K images daily for automation efforts.

Cedar Rapids, Iowa Area

- Contributing to the full development life cycle of machine learning solutions for various million-dollar aerospace contracts. - Developing a highly scalable and containerized cloud-based backend infrastructure that serves the FAA’s weather camera data collection and analysis system; efficiently processes about 100K images daily- Building performant ETL data pipelines and frontend interfaces for end-to-end machine learning and analytics systems- Writing technical specifications based on stated business and technical requirements for projects in several aviation-specific domains (weather forecasting, cabin behavior, pilot health, software requirements, etc.)- Consulting with clients and subject matter experts to prototype, A/B test, refine, and debug programs to meet needs- Leading the ML/AI Community of Practice (CoP) – the largest active CoP with 1000+ members and 300+ monthly attendees - Establishing connections and aligning goals of ML/AI leaders across Collins Aerospace, Raytheon, and UTC

United States

We pull together and mentor cross-functional teams of students and career-changers to build meaningful side-projects

United States

- Leading a fellowship program for beginners to learn by building end-to-end team-based machine learning (ML) applications. We hold weekly meetings to discuss progress and set action items (participants from across 4 continents and 9 countries)- Managing a discord server of over 500 ML practitioners for meaningful discussions on relevant ML research and applications- Producing videos on applied ML advice and guidelines along with videos covering personal ML side-projects applied to unique problem spaces (www.youtube.com/chroniccoder, 60K+ views, 250K+ watch minutes)

- Developed an interpretable content-based anime recommendation API that is integrated with a popular anime recommendation website (www.randomanime.org) averaging 75K unique users monthly- Implemented an extensive anime knowledge graph with state-of-the-art NLP techniques using gensim and networkx- Established a scalable data ingestion pipeline from multiple anime-specific data sources- Utilized FastAPI, Docker, AWS EC2, and REACT to create an efficient and portable prototype web application and API

- Managed a university residence hall floor of 51 ethnically and academically diverse undergraduate students- Assigned floor with statistics and data science learning community students where we discussed and addressed many school, career and personal inquiries- Confronted and resolved issues on diversity, racism, alcohol abuse, academic performance and social acceptance- Planned and executed weekly floor events based on residents' interests and self-development goals

Cedar Rapids, Iowa Area

- Implemented multivariate and multi-step LSTM recurrent neural networks to predict student’s cognitive workload- Engineered a Flask web application to help student pilots measure their performance individually and against their peers as well as help instructors measure their classes’ progress - Developed interactive and dynamic data visualizations of the practical data with D3.js - Configured a Flask RESTful API to effectively handle get and post requests - Collaborated with my manager to produce Docker images to efficiently run the whole system through containers(Tools: Flask, Jinja, D3.js, Bootstrap, JQuery, Pandas, Spark, Cassandra, Docker, Jupyter Notebook, Keras, Tensorflow | Languages: Python, Javascript, HTML, CSS)

Lafayette, Indiana Area

- Assisting two CS180 (Java Programming) classes of around 24 students; guiding and encouraging them in their lab assignments each week- Composing test cases for a large programming project to be used for the grading of over 360 students (Tools: Junit; Languages: Java)

Lafayette, Indiana Area

- Creating a project website and host web servers running real time queries on the Tornado big data system- Inventing a proper format for open-sourcing the code of Tornado (Tools: Apache Storm, Kafka, Apache Zookeeper, Maven; Languages: Java, HTML)

Lafayette, Indiana Area

- Independently developing a Java Syntax game to help CS 180 (Java Programming) students to learn and understand the use of Java functions through the puzzle solving mechanics and procedurally generated gameplay- Created Java parser and user interface to change objects in the game world in a variety of ways based on String function calls by the user- Leading a team of CS180 students in creating a series of conceptual games to help future students understand Java syntax or concepts- https://github.com/chriskok/StringOfMurders (Tools: Unity3D, Maya, Adobe Photoshop; Languages: C#)

Doctor Of Philosophy - Phd, Human Computer Interaction

Bachelor'S Degree, Computer Science, Cgpa: 3.97

Graduated with Highest Distinction. Fun fact, I've held 5 student jobs while at Purdue and every abbreviation ends with an A: Student.