I build education systems that diagnose where a learner’s foundation begins to break, trace the prerequisites beneath the gap, and turn that evidence into a personalized path forward.
I’m the founder of Foundations to Mastery, where I work across product strategy, software engineering, curriculum architecture, diagnostic design, data systems, and educator operations. The goal is not simply to deliver more content. It is to make each instructional decision explainable: what the learner knows, what remains uncertain, where the earliest meaningful gap begins, and what should happen next.
Most learning platforms start with content. I start with evidence.
| System | What it is designed to do |
|---|---|
| Diagnostic intelligence | Identify where a learner’s foundation breaks without reducing the learner to one grade-level label or hiding unmeasured gaps. |
| Curriculum infrastructure | Turn standards, lessons, practice, figures, assessments, and teacher materials into structured, versioned learning systems. |
| Personalized learning paths | Convert diagnostic findings into sequenced instructional cycles with clear prerequisites, checkpoints, and reassessment. |
| Tutor and operations tools | Give educators the evidence, materials, student context, and next actions needed to run a focused session without rebuilding the system by hand. |
| Educator-reviewed AI workflows | Use AI for grading support, synthesis, and drafting while keeping consequential decisions traceable and subject to human review. |
Intake → Adaptive Diagnostic → Evidence Map → Educator Review → Personalized Path → Continuous Reassessment
- Diagnose before teaching. A polished lesson is not useful if it starts above the point where the learner’s understanding first fractured.
- Every conclusion should have a trace. A finding should be connected to observable evidence, its assumptions, and the next action it supports.
- AI proposes; educators confirm. Automation should remove repetitive work without disguising uncertainty or removing professional judgment.
- Build infrastructure, not one-off artifacts. The strongest curriculum, assessment, and tutoring workflows are reusable, testable, and maintainable.
- A learner is not a grade label. Strengths and gaps can coexist across strands, representations, and years of prerequisite knowledge.
- Program Design and Abstraction — coursework and projects in program structure, abstraction, and computational problem solving.
- Data Science Portfolio — applied machine learning, statistical modeling, inference, and data analysis.
- ChatGPT Flashcards to Anki Converter — a study-workflow tool that converts structured AI-generated flashcards into an Anki-ready format.
- Berkeley Projects — Java, data structures, graph traversal, procedural generation, and software design projects.
Product and application engineering: React · JavaScript / TypeScript · Python · Supabase · PostgreSQL · Deno · REST APIs
Learning infrastructure: diagnostic systems · prerequisite graphs · curriculum modeling · assessment design · progress instrumentation · tutoring workflows
Content and research tooling: Markdown · LaTeX · KaTeX · deterministic build pipelines · data analysis · reproducible documentation
Most Foundations to Mastery product code and all student information remain private. Public work is shared with clear boundaries around privacy, provenance, reproducibility, and what the evidence can actually support.