Product thinker and hands-on AI builder with a background in technical strategy and data analysis.
I’m interested in how AI and data can help people make better decisions. I build prototypes to explore that question—from defining the user’s problem and shaping the experience to testing the system and understanding its limitations.
I use AI-assisted development to turn ideas into working products and demos, with a focus on useful experiences, clear trade-offs, and honest evaluation. Here, I share what I build, the decisions behind it, and what I learn along the way.
🔎 Frame the problem → 🛠️ Build & test → 📊 Evaluate → ✍️ Share what I learn
Helping Georgia students explore scholarships, understand potential matches, and plan their next steps. Combines cited AI explanations with explicit eligibility checks and visible uncertainty.
Product focus: Guided discovery · Trust · RAG evaluation
Course demo with a limited corpus; not a final eligibility determination.
Live demo · Case study · Code & product decisions
Helping hosts turn recipes, menu choices and available time into a manageable prep plan. Suggests recipe swaps or bought-food alternatives, checks whether the revised menu fits, and leaves approval with the host.
Product focus: Agentic planning · Human approval · Evaluation
Team 44 course project using LangGraph and LangSmith. Runs locally; calendar booking requires approval.
Code & product decisions · Watch demo
Helping young readers practice with short stories, visible progress, and a way to flag answers for review. Built from a family reading-practice project and adapted into a public learning edition.
Product focus: Learning experience · Fair answer feedback · Human review
Local-only demo with original sample stories and synthetic learner data; the family app remains private.
Code & walkthrough · Case study
Helping event operations leads spot attendee friction and decide where to investigate. Brings queues, occupancy, feedback, and support signals into an interactive dashboard.
Product focus: Decision support · Data storytelling · Human judgment
Synthetic data and simulated replay; no LLM or automated dispatch.
Exploring: AI evaluation, retrieval-augmented generation, and tools that help people make better decisions.