Learning Experience Design | AI-Enabled Training Systems
Designing learning systems that improve performance, not just completion.
I design AI-enabled, evidence-based learning experiences for complex topics so teams can
learn faster, perform better, and scale quality.
30-Second Recruiter Summary
Belle Li is a Learning Experience Designer and Purdue Ph.D. candidate who builds
performance-focused learning systems across startup product training, graduate education,
and AI-enabled workflows. She has led cross-functional curriculum operations,
co-instructed graduate courses, produced scalable learner-support assets, and translated
research into implementation-ready playbooks. Her work is strongest where complexity is
high: identifying learner friction, designing targeted interventions, and improving transfer
through evidence-based iteration.
AI-Enabled Learning Design
PeteChat: Guardrailed AI Tutor for Higher Education
Designed an assessment-aware AI learning assistant at Purdue — built to scaffold thinking, not deliver answers. $84,787 funded, 4 iterative phases, deployed to undergraduate courses.
Open case
Course & Program Design
Graduate Course Redesign + 13 Support Videos
Redesigned Purdue EDCI 557 for asynchronous graduate learners with scaffolded
assignments, rubrics, and weekly video guidance.
Open case
Product / Training Operations
LMS + Content Operations at Startup Scale
Co-led HiLink LMS direction and curriculum operations with a coordinated team model and
300+ learning materials delivered.
Open case
About
I’m Belle Li, a Learning Experience Designer and Ph.D. candidate in Learning Design and
Technology at Purdue University. I build learning experiences where clarity, transfer, and
measurable outcomes matter.
Across university teaching, startup product work, and research collaborations, I design
curriculum systems, learner-support assets, and AI-supported workflows for multilingual,
non-traditional, and professional learners.
My approach is practical: identify where learners get stuck, build targeted support, then
evaluate and iterate.