T127: Learning Design, Technology & Innovation
Harvard Graduate School of Education — Spring 2026
A semester-long studio bridging learning theory with real-world design practice. I worked with Harvard's Teaching and Learning Lab (TLL) on a university-wide Data Fluency initiative — a foundational program to help staff across all Harvard schools build practical data literacy skills.
The course began with deep exploration of learning theories — Cognitivism, Constructivism, and Connectivism — and their application to adult learners. Through extensive stakeholder engagement and discovery work, I learned that the problems Harvard staff actually faced were very different from what we initially assumed. This gap between assumed and surfaced problems became the defining lesson of the semester.
The portfolio records how I thought about the data fluency course and how my thinking moved as I learned new concepts. In the spirit of visible thinking, it serves as a place to showcase the artifacts I make and the reasoning behind them — legible to me and to anyone learning alongside me.
Capstone
T127 Course Capstone: Generative UI for Education
completedA proof-of-concept generative interface that scaffolds personalized learning modules for Harvard's data fluency course. The system adapts to each user: rendering structured support (worked examples, guided steps, glossary panels) for novices and progressively stripping it away as the learner model indicates competence.
User answers to dynamic questions help build a Harvard-wide data dictionary through learnersourcing. This creates a richer content base that can be cross-referenced by learners in the future. Quality control is managed through a tight, inspectable, open-source ingestion pipeline.
Semester Journey
-
T127 Reflection 3
Capstone reflection on designing a Generative UI system for Harvard's data fluency course.
Supporting Artifacts
- Analysis T127 Capstone Analysis
- Docs T127 Docs (mid-course)
Key Takeaways
- Discovery before design: Actual user data beats assumed problems. The five Ds aren't equally weighted, and discovery is where I was cutting corners.
- From solutionist to designer: Sit with the problem longer than feels comfortable. AI is not always the answer.
- Human-in-the-loop: Build around learning designers, not just learners. Every part of the codebase should be calibratable by them.
- Visible thinking: Portfolios as source of truth for reasoning — legible to self and collaborators.
- Context matters: Data fluency looks different for a GSAS admissions officer than a GSE career consultant.