Autumn Clark
Download CV

Teaching

Belonging first, then rigor

Students are better able to undertake difficult intellectual work when they believe they belong to a community of learners and that their instructor is invested in their success.

One of the most memorable moments of my undergraduate education was when a professor I had never met welcomed me by name, held the door for me, and asked several students — whom he also addressed by name — to make room for me in their row.

His actions communicated that I was not anonymous, that my presence mattered, and that my success was important to him.

From my teaching statement

Care is not shielding

Caring for students does not mean shielding them from struggle. It means building the self-efficacy they need to meet rigorous expectations — the confidence that comes from facing and overcoming problems.

Start from what students know

Technically complex material becomes approachable by beginning with what students already understand, then progressively revealing the underlying computational problem.

Practice before judgment

Students gain confidence when they can apply unfamiliar concepts, receive feedback before a final evaluation, and revise their thinking rather than being judged on a first attempt.

Signature lesson

Is there a cat in this image?

A guest lecture on computer vision and convolutional neural networks (CNNs) for Artificial Intelligence, Machine Learning, and Deep Learning in Business Information Technology at Virginia Tech. This is not a CNN tutorial. It is the structure of the lesson: familiar experience, then a conceptual model, then technical detail, then authentic application.

  1. 1 · Familiar experience

    Does this image contain a cat?

    The lecture opened with a task students regarded as almost trivial: decide whether each of several images contained a cat.

  2. 2 · Familiar experience

    How do you know?

    Then students were asked to explain how they knew. Their difficulty articulating a judgment they make instantly was the point.

  3. 3 · Conceptual model

    The computational problem

    That difficulty illustrates why teaching a computer to interpret visual information is hard: the computer receives only a grid of numbers.

  4. 4 · Conceptual model

    Borrowing from human vision

    Human visual processing became the intuitive bridge to receptive fields — small regions that each respond to local patterns — and to hierarchical feature extraction, from edges to parts to whole objects.

  5. 5 · Conceptual model

    One section at a time

    In an interactive activity, images were revealed one small section at a time. Students identified the image and named the features that informed — or misled — their judgments.

  6. 6 · Technical detail

    Kernels, pooling, and feature maps

    Only after that conceptual foundation were the mechanics introduced: kernels sliding across an image, the feature maps they produce, and pooling that condenses them.

  7. 7 · Authentic application

    Why it matters

    The lecture closed by connecting image classification to real-world applications, particularly in healthcare.

  8. 8 · Throughout

    Checking understanding along the way

    Questions embedded throughout the lecture made it possible to assess understanding before moving forward, rather than discovering confusion at the end.

Teaching artifacts

Lessons and materials

Use the arrow buttons or arrow keys to move through slides. Slides never advance on their own.

Computer vision and convolutional neural networks

Audience
Students in Artificial Intelligence, Machine Learning, and Deep Learning in BIT, Virginia Tech (guest lecture, delivered twice; Spring 2026 version shown)
Learning objective
Understand the capabilities, limitations, and real-world applications of CNNs and image classification, and explain a technical diagram of a CNN model.
Teaching challenge
Students may at first feel intimidated by complex technical material.
Instructional choice
Begin with a familiar task (is there a cat?), build a conceptual model from human vision, reveal images one receptive field at a time, and only then introduce kernels, pooling, and feature maps.
Evidence or reflection
The instructor invited the lecture back the next time the course was offered. After the second delivery, one student reported their understanding of CNNs had risen “from a 4 to a 7.”
Slide asking, “Is there a cat in the following pictures?”
Slide with photos of cats and a puppy, asking, “How did you know that?”
Slide titled “You just do,” explaining that the brain labels visual input with meaning before we are conscious of it.
Diagram of the visual cortex: simple features such as lines and curves combine into shapes and, finally, a recognized object.
Instructions for a learning activity: images will be revealed one receptive field at a time; type in the chat when you know what you are looking at.
A fully revealed image from the activity: the Virginia Tech HokieBird mascot on a football field.
Slide connecting the activity to how AI classifies images, one small group of pixels at a time.
Worked example of convolution: a 3×3 filter slides across an input grid of numbers to produce an output feature map.
Max-pooling example reducing a 4×4 grid of numbers to a 2×2 output.
Application slide on healthcare showing brain scans with segmented regions.
Homework slide: comment your own CNN, then compare the performance of several CNN architectures.

Slide 1 of 11

Preparing to teach: a Ph.D. seminar session

Audience
Doctoral students in the Ph.D. Seminar in AI, Machine Learning, and Deep Learning, Virginia Tech (Fall 2026), each preparing a 45-minute presentation and demonstration of an AI technology
Learning objective
Plan a 45-minute lesson around what students should understand or be able to do by the end, using evidence-based active-learning strategies.
Teaching challenge
Forty-five minutes is not enough time to teach everything you know about a subject, and the default lecture model encourages passive learning.
Instructional choice
Modeled the approach rather than describing it: a think-pair-share on memorable classes, a ranking activity on the passive–active spectrum, and the CNN lecture as a worked case study.
Evidence or reflection
Invited by the course instructor to lead the session. It opened and closed with brief participant surveys.
Opening activity: think of a lesson that stuck with you and share with the person next to you.
Lesson planning step one: ask what students should understand or be able to do at the end of 45 minutes.
Definition of active learning beside Bloom’s taxonomy, with examples such as polls, case studies, and think-pair-share.
A spectrum from passive learning (mostly professor talking) to active learning (high student effort).
Teaching formats, from non-interactive lecture to student-led projects, arranged along the passive–active spectrum.
Slide titled “Passive != Bad,” explaining that foundational knowledge still requires direct instruction.
Summary slide: explain something, have students think or do something with it, then continue and repeat.
Table of Autumn’s strategy for teaching technical material, applied to the CNN lecture: learning goal, familiar starting point, conceptual model, technical detail, application.
Autumn’s practical rules of teaching, such as students engaging more when they can tell their teacher cares about their learning.

Slide 1 of 9

Courses

What I am prepared to teach — and why

When teaching emerging technologies such as AI, I integrate technical skills with organizational relevance: for business students, an AI course can combine technical foundations with organizational applications, ethical analysis, and hands-on evaluation of AI systems.

  • AI, machine learning, and deep learning in business

    • Guest lecturer, AI/ML/DL in Business Information Technology (Fall 2024; Spring 2026)
    • Teaching assistant for the course (Fall 2024) and for the Ph.D. seminar in AI, ML, and DL (Fall 2024; Fall 2026)
    • Built a CNN-based research artifact; works in PyTorch, Keras, and Azure ML Studio

  • Cybersecurity management

    • Teaching assistant, Cybersecurity Management I (Fall 2024)
    • Research program in healthcare privacy and information disclosure

  • Data analytics and visualization

    • Teaching assistant, Advanced Excel — approximately 120 students (Fall 2020)
    • R, Python, SQL, Stata, Tableau, and Power BI in research practice
    • Coursework in data analytics

  • Systems analysis, web development, and technical IS

    • Founder and lead developer of OkWellThen, a healthcare price-transparency platform
    • Solutions architect intern, Adobe (Summers 2019 and 2020)
    • JavaScript, Vue.js, Node.js, MongoDB, HTML/CSS, C++, and C#

  • Introductory and managerial information systems

    • B.S. in Information Systems and Master of Information Systems Management (Healthcare Track)
    • Industry experience at Adobe, Sorenson Impact Foundation, and the BYU Healthcare Initiative

  • Project management

    • Coordinated teaching assistants at Virginia Tech (Fall 2024)
    • Coordinated a three-month pilot mentorship program pairing 15 students with industry professionals
    • Coursework in project management

  • Research methods (doctoral)

    • Led a Ph.D. seminar session on preparing to teach with evidence-based, active-learning principles
    • Advised doctoral students on research proposals and evaluated weekly reading syntheses
    • Qualitative analysis, factorial survey and laboratory experiments, SEM, hierarchical modeling, and scale development

Experience

Teaching and mentoring

  1. Fall 2024; Spring 2026

    Guest Lecturer

    Artificial Intelligence, Machine Learning, and Deep Learning in BIT · Virginia Tech

    Interactive lectures on workplace applications of generative AI and on computer vision and convolutional neural networks.

  2. Fall 2024

    Teaching Assistant (TA Coordination)

    AI, ML, and DL in BIT; Cybersecurity Management I · Virginia Tech

    Coordinated other teaching assistants, assisted with course administration and grading, and delivered guest instruction.

  3. Fall 2024; Fall 2026

    Teaching Assistant

    Ph.D. Seminar in AI, Machine Learning, and Deep Learning · Virginia Tech

    Invited to lead a session on preparing to teach; held office hours on research proposals and evaluated weekly reading syntheses.

  4. Fall 2020

    Teaching Assistant

    Advanced Excel · Brigham Young University

    Mentored approximately 120 students; led midterm and final reviews on goal seeking, financial calculations, and VBA; gave formative feedback on approximately 40 final projects.

  5. Spring 2016

    Teaching Assistant

    Interdisciplinary Honors Course · Brigham Young University

    Tutored 50 honors students in introductory physics and mentored interdisciplinary capstone papers.

Development

Pedagogical training

  • Completed

    Introduction to Evidence-Based Undergraduate STEM Teaching

    CIRTL

  • Completed

    Incorporating Active Learning

    CIRTL

  • Completed

    Difficult Dialogues in the Classroom Workshop

    CIRTL

  • In progress · expected 2026

    CIRTL Associate Teaching Program

    Center for the Integration of Research, Teaching and Learning

Practices I will bring to my courses

  • Start-of-course survey

    A brief survey on students’ backgrounds, interests, prior experience, professional goals, and concerns about the subject — answered individually in small courses and used to shape examples, activities, and outreach in large ones.

  • Informal midsemester feedback

    Students identify what is supporting or impeding their learning while there is still time to respond.

  • Practice before evaluation

    Opportunities to apply unfamiliar concepts, receive feedback before a final evaluation, and revise their thinking.

The full teaching statement is available with other application materials.

Teaching statement (PDF)CV (PDF)