Daniel Umana, MSEd.

I build the tools that make AI literacy real.

Curricula, frameworks, and working software for faculty and students.

Digital Learning Center Manager at Montgomery College. Ed.D. candidate in Education Policy and Leadership at American University.

Daniel Umana, wearing glasses and a gray quarter-zip, lit by a teal spotlight.
Montgomery College. American University.

The work

Every one of these was built
for a specific course, workshop, or cohort.

What follows is each build as it actually looks, what it does, and who asked for it. Three of them are gated, because they hold a cohort's work or run inside someone else's course. Those are shown here and not linked.

10 built 6 public 3 invite or cohort only 1 program

AI Literacy Tools 5 builds

01

ALIGN

AI Levels for Instructional Guidance and Navigation

Six levels that give faculty and students shared language for when and how AI belongs in coursework.

  1. 0 No AI Use Independent work. No generative AI.
  2. 1 Limited AI edits clarity after the student drafts.
  3. 2 Brainstorming AI plans and generates ideas only.
  4. 3 Collaborative Creation AI drafts parts. The student revises visibly.
  5. 4 Advanced Integration Multi-step work. The student judges accuracy and bias.
  6. 5 Innovation and Exploration The student designs the AI workflow.
The ALIGN framework site, headed "ALIGN, AI Levels for Instructional Guidance and Navigation", set in Montgomery College purple.
mc-align.vercel.app, the public framework site.
  • Six assignment-level categories, from no AI use to full integration
  • A communication tool, not a mandate, a ranking, or a detector

The companion scoring app is invite-gated for faculty.

Two versions of one essay prompt side by side, each sentence shaded from AI-familiar to human-specific, scored 82 out of 100 and 21 out of 100.
Sentence scoring, original against redesign.

02

Epistemic Density Tool

Scores every sentence of an assignment by how AI-familiar versus human-specific it is, then suggests redesigns that protect original student thinking.

  • Sentence-level color coding on a human-versus-AI density spectrum
  • Live heat map, semantic map, and an assignment-level density gauge
  • Rewrites compared side by side, exported as PDF or Word
Live, invite only

It is in use and it is not public, so there is no link here.

reKWL asking a reader to ground a claim in the text, with the Know, Want to know and Learned steps above a four-stop reading level picker.
rekwl.vercel.app, with the reading level picker.

03

reKWL

A Socratic reading chart that never fills itself in. Students write what they know and what they wonder; reKWL answers with questions tuned to their reading level.

  • Know What they already bring.
  • Wonder What they want to find out.
  • Learned What they took from the text.

reKWL writes none of it. It answers with a question, pitched at one of four reading levels, from early elementary through higher ed.

Runs on Llama 3.3 through Groq, streamed from a server route.

04

AI Policy Builder

EDU 280, Social Justice and Urban Education, American University

You make the policy.

The district lives with it.

Results come back by group.

Then you decide again.

One decision leaves the room. It does not arrive the same way for every group of students. The simulation reports back disaggregated, and you keep playing.

AI Policy Builder: a darkened legislative chamber behind the line "Run AI policy for a school district. Live with what it does," with a red Take office button.
edu280-policy-builder.vercel.app, the opening chamber.

A research-grounded policy simulation. The player runs AI policy for a fictional district and lives with what it does, reported back disaggregated by student group.

  • Four rounds across three school years, 2026 to 2029
  • Four bands playable: K-5, 6-8, 9-12, and higher ed
  • 207 cited sources across four verified evidence bases
  • One deterministic engine. No accounts, no database, no analytics

Built for Dr. William N. Thomas IV.

Cover of the open-access curriculum "AI Literacy for Career and College Success" by Daniel Umana, M.S.Ed.
The open-access curriculum, cover.

05

AI Literacy Microcredential

A stackable AI literacy credential with an open-access curriculum, designed and run end to end at Montgomery College’s Digital Learning Center.

  • End to end: intake, registration, tracking, evaluation, reporting
  • Open-access curriculum, written by Daniel
  • Foundational, responsible AI skills for community college students
Program, Montgomery College

Workshop Companions 2 builds

AI That Works for You: a sixty minute thinking workshop page reading "Use AI to improve the process around your thinking. Not to replace it."
ai-that-works-for-you.vercel.app, the session home.

06

AI That Works for You

Built for an MCRPA session on practical AI. The workshop’s rule, guardrails, and practice stay where participants can return to them.

The rule, on screen all session
Use AI to improve the process around your thinking. Not to replace it.
  • Guardrails and practice prompts, still there the week after

07

Practical AI for Small Business

A hands-on 90 minute workshop companion for small-business owners: use AI to improve the process around your work, not to replace your judgment.

Practical AI for Small Business: a hands-on ninety minute workshop companion page for small business owners.
practical-ai-small-business.vercel.app, the session home.
  • Ninety minutes, built to be worked through, not watched
  • Written for owners and operators, not for a classroom

Doctoral Work 3 builds

08

PDSA Outline

Coaches a Plan, Do, Study, Act improvement outline phase by phase, then exports a submission-ready Word document.

  1. Plan Name the problem of practice and the change you intend to test.
  2. Do Run the change small, on purpose, and write down what happened.
  3. Study Look at what came back, including the part you did not expect.
  4. Act Decide. Adopt it, adapt it, or abandon it, and say why.
The Plan-Do-Study-Act cycle as the outline draws it, four coloured arcs around a ring, titled for environmental scans and needs assessments in EDU 703.
The outline’s opening card.
  • A per-phase coach that asks questions before it suggests anything
  • Antiracist guardrails written into the outline itself
  • It draws out your thinking. It will not write your Problem of Practice for you
Live, cohort passcode

It runs for the cohort it was built for, so there is no link here.

09

DoP Self-Assessment

Replaces a paper competency self-assessment with private, immediate feedback and two generated PDFs.

DoP Competency Self-Assessment: nine competencies in three categories, with a radar chart of one person’s results.
dop-self-assessment.vercel.app, the results view.
Competencies
9
Sub-elements
33
Generated PDFs
2
Scoring
Deterministic
Accounts
None
Server storage
None
  • Tailored reflection prompts, written from the answers you gave

10

Abbott SDT Research App

Helps chapter authors of a forthcoming book find the episodes, rationale, and citations that support a claim.

From a claim to a citation

  1. 01 A claim from your chapter
  2. 02 The coded episodes that support it
  3. 03 An APA-7 citation built from the episode credits
  • Chapter-first browse plus BM25 search over coded episodes
  • A draft analyzer grounded in Self-Determination Theory
Live, invite only

It holds a book team’s working notes, so there is no link here.

The Abbott SDT research app: a "find the episodes for your chapter" search panel, with an empty search field and four example topics beneath it.
The search panel.
Browse by chapter: seven chapter cards, each counting the episodes coded Primary and Secondary for that chapter.
Chapter browse, with each chapter’s episode counts.

Research

From Compliance to Connection

A live workspace for AI-personalized assignment design in doctoral education.

Type
Design case
Site
American University
Ran
June 2026

Most institutional responses to student AI use, detection and policy, sit downstream of the behavior they target. This design case takes the opposite approach. A free, browser-based workspace personalizes a doctoral writing assignment around each student's own experience before the writing begins.

The AI designs the prompt. It never writes the response.

Detection sits downstream.

01 Assignment written
02 Student sits down
03 Writing happens
04 Work submitted
05 Detection and policy

Steps 01 to 02

Where this workspace acts. Before the writing begins.

Step 03

The writing itself. Neither side acts here.

Steps 04 to 05

Where most institutional responses land. After the behavior they target.

Negentropy, after Schrodinger

Compliance

One prompt for everyone. Effort scatters into producing the artifact, and the fastest route through it is a model that has never met the student.

Connection

Route the prompt through what the student carries and the motivational energy organizes around the work. A student writing about their own seventh-grade classroom has a reason to do the work themselves.

  • Self-Determination Theory. Ryan & Deci, 2017
  • Culturally responsive teaching. Hammond, 2015

Effort scatters. One prompt, every direction at once.

Effort organizes. The prompt routes through the student.

Illustration. It encodes no data.

How it works

The assignment object as it assembles. The rubric is present from the first screen and never changes.

  1. 01

    Frame

    The assignment's fixed objectives stay visible before any writing happens, and the rubric never changes from student to student.

    Objectives fixed. Rubric unchanged.

  2. 02

    Listen

    Four short intake questions collect the student's context: settings and roles, one critical moment, their current inquiry, and the lens it created.

    Four questions. Nothing stored.

  3. 03

    Rewrite

    The model generates a personalized version of the prompt from that intake. It may quote the student's own phrasing. It may not invent details.

    The AI designs the prompt. It never writes the response.

  4. 04

    Hand off

    The student downloads a Word document with the personalized prompt, their intake, the unchanged rubric, and a writing space.

    The AI is never called during the writing.

What actually happened

It ran with a real cohort. It claims no outcomes.

Where
An American University Ed.D. course, June 2026. It ran during the term, on coursework that counted.
Consent
Cohort members opted in and used it to complete a graded positionality memo, with the instructor's prior awareness. Daniel wrote his own memo inside it.
Data
No login, no database, no analytics. It stores nothing.
Claims
The outcome questions are unmeasured by design. The paper specifies the pilot study that comes next.

Public, free, no account. A full cycle, landing to download, takes five to ten minutes.

Writing and speaking

Ideas in public

Long-form writing on AI literacy, pedagogy, and equity in education, and the talks and workshops where the same questions get asked out loud.

Writing

6 articles, published on LinkedIn

  1. Critical pedagogy Freire, the Banker, and the Chatbot Read on LinkedIn (opens in a new tab) Most AI-in-education debates are two camps shouting past each other. Paulo Freire is the missing lens: his critique of "banking" education reframes what AI literacy should actually be doing for students.
  2. AI literacy Green Means Go: Accelerating AI Literacy in Education Read on LinkedIn (opens in a new tab) Students cannot be expected to think critically about tools they were never taught to use, it is like expecting someone to drive safely without ever being shown how. A case for moving deliberately, and faster, on AI literacy.
  3. Frameworks AI, Pedagogy, and a Framework Worth Revisiting: Why TPACK Still Matters Read on LinkedIn (opens in a new tab) Generative AI did not break pedagogy; it raised the stakes. TPACK still helps us make sense of where technology, content, and teaching actually meet.
  4. Equity Deficit Thinking in the AI Era: An Invitation to Reflect Read on LinkedIn (opens in a new tab) Every few months a headline predicts higher education will collapse because of AI, or that students will stop thinking. That is deficit thinking, and an invitation to reflect on the story we tell about students.
  5. Digital equity AI Literacy Is the New Digital Divide Read on LinkedIn (opens in a new tab) AI is among the most influential learning tools of our time, and students are already meeting it everywhere. If educators look away, the divide only widens. Why we cannot afford to.
  6. Learning Generative Love and Generative AI Read on LinkedIn (opens in a new tab) Generative love is not a phrase we usually hear next to artificial intelligence. But it offers a different, more humane way to think about what learning can be in the age of AI.

Speaking

Conferences, campus sessions, and ongoing workshops

Daniel Umana speaking beside a projected slide about generative AI to a classroom of seated students at Montgomery College.
Leading AI Literacy workshops at Montgomery College.
Recorded talk, hosted on YouTube.

Featured talk

AI: Friend or Foe?

Future of Tech Lunch & Learn · Montgomery College ignITe Hub

May 2026

The video loads from YouTube only when you press play, and it plays with sound.

  • Building Student-Facing GenAI Resources CALD Conference 2025, Rockville, MD Speaker 2025
  • Maryland Latinos Unidos Futurist Summit 2025 Maryland Latinos Unidos Presenter 2025
  • AI literacy workshops for students, faculty, and K-12 audiences “AI: Friend or Foe?”, “AI Study Ready”, “AI and You”, and more Designer & Facilitator Ongoing
  • Montgomery College AI Club Mentorship Mentor 2024 - Present

Honor

Joyce Gray Memorial Award

Peer-selected award, Montgomery College Leadership Development Institute (32nd cohort), 2026.

Experience and credentials

From the classroom to AI literacy leadership.

Nearly a decade in K-12 as a teacher, grade chair, school culture leader, and founding dean, before higher education. 6 roles, in the order they happened.

  1. K-12
  2. 01 Jun 2017 - Jul 2021

    Dean of Students / Grade Chair & 4th Grade Teacher

    Breakthrough Public SchoolsCleveland, OH

    Led the rapid transition to virtual operations during COVID-19.

    • Delivered grade-level curriculum and trained staff on instructional technology.
  3. 02 Jun 2021 - Jul 2022

    Dean of Students

    DC Scholars Public Charter SchoolWashington, DC

    Led organization-wide data collection and analysis informing student support strategy.

    • Planned, budgeted, and managed a student-facing operation with full operational oversight.
  4. 03 Jun 2022 - Jul 2023

    School Culture Specialist

    Rocketship Public SchoolsWashington, DC

    Ran data-informed operational plans supporting daily school functions.

    • Designed and facilitated professional development that strengthened consistency of practice.
  5. 04 Jun 2023 - Jul 2024

    Founding Dean of Students

    Uncommon Public SchoolsBrooklyn, NY

    Co-developed the mission, vision, and values for a new campus and the operating systems behind them.

    • Designed and ran data-driven student support systems and iterated on interventions.
  6. Higher education
  7. 05 Aug 2024 - Present

    Digital Learning Center (DLC) Manager

    Montgomery CollegeRockville, MD

    Leads the team, budget, and operations of a centralized AI literacy and digital learning hub, including a technology lending program.

    • Led the cross-functional faculty-and-staff group that created the college-wide ALIGN framework; co-authored its white paper and faculty guides.
    • Launched a credentialed AI literacy microcredential and authored the open-access curriculum behind it.
    • Supervises and develops staff and partners with faculty across departments to grow AI literacy participation.
  8. 06 May 2026 - Present

    Graduate Research Assistant

    American University, School of EducationWashington, DC

    Supports Dr. William N. Thomas IV, Ed.D., on a qualitative study connecting AI-enabled assignments and Self-Determination Theory.

    • Leads the coding workflows, codebook iteration, and research documentation toward a forthcoming book proposal.
    • Designed and built the AI-powered research application behind the work (syllabus decoding, assignment redesign, database search).

Credentials

Education

  • Ed.D., Education Policy & Leadership

    American University, School of Education

    Expected May 2028

  • M.S.Ed.

    Johns Hopkins University

    2019

  • B.S., Public Health Science

    University of Maryland

    2016

Certifications

  • AI+ Educator

    AI Certs

    2025

  • Leading Responsible AI in Organizations

    LinkedIn Learning

    2025

  • Generative AI Leadership & Strategy

    Vanderbilt

    2024

  • Generative AI for Educators & Teachers

    Vanderbilt

    2024

  • AI in Education

    University of Pennsylvania

    2024

  • Generative AI: Impact, Considerations & Ethical Issues

    IBM

    2024

Skills

  • AI & GenAI (ChatGPT, Copilot, Gemini, Claude incl. Claude API)
  • App building with Claude Code
  • Next.js / React, Python / Flask, Supabase, HTML/CSS/JS
  • Curriculum & microcredential design
  • Program management & data reporting
  • Pressbooks, H5P, Microsoft 365, Google Workspace
  • Bilingual English / Spanish

About

Classroom first, then the tools.

I came up through the classroom. I spent nearly a decade in K-12 before I brought that experience to higher education.

Today I manage Montgomery College's Digital Learning Center, where I lead the team and the operations behind a college-wide AI literacy effort. I launched a credentialed AI literacy microcredential and wrote the open curriculum behind it, and I led the cross-functional faculty-and-staff group that created the ALIGN framework.

I am an Ed.D. candidate in Education Policy and Leadership at American University, where I also serve as a Graduate Research Assistant. I am a first-generation college graduate and bilingual in English and Spanish.

None of these tools will do a student's thinking for them.

That rule came out of the classroom, not out of a policy document.

Contact

Get in touch.

Workshops, AI literacy work with faculty or students, and questions about the research above are all good reasons to write.