isa
$
$
$ currently building adaptive STEM tools + audio software
Technologies
Experience
- Incoming Associate Web Developer role focused on building and maintaining server-side systems, APIs, and data-backed application infrastructure.
May 2026 – Present
- Architected and shipped an AI-adaptive learning engine in Flutter/Dart using Bayesian Knowledge Tracing, modeling each student's latent knowledge state in real time to dynamically adjust explanation depth, code visibility, and hints across lessons.
- Engineered a multi-signal delivery layer combining 4 inputs (quiz performance, time-on-page, hint reliance, and per-page mistake counts) into a continuous difficulty score, improving adaptation accuracy over a quiz-only baseline.
- Designed the persistence layer on Supabase (Postgres) to carry adaptive state across sessions and recover partial signal on mid-lesson exits, eliminating dropped engagement data that previously reset on every navigation.
May – July 2026
- Fixed a React scroll-listener bug with a shared requestAnimationFrame-throttled hook, lifting the homepage Lighthouse Performance score from 73 to 88.
- Cut image payload 88% (105.9MB to 12.7MB) across an 11-page Next.js site, dropping Largest Contentful Paint up to 38% · validated with controlled before/after Lighthouse audits.
- Root-caused 5+ rendering bugs (mobile Safari scroll jank, competing animations, layout thrashing), cutting Total Blocking Time by up to 83%.
- Built a 6-page production site from scratch (HTML/CSS/JS) and modernized a second live client site across 5+ pages.
May 2024 – May 2025
- Engineered a two-sided marketplace platform used in behavioral-economics experiments with 4,000+ users and LLM-based agentic AI sellers using React, Node.js, and JavaScript.
- Designed backend REST APIs and refactored server-side logic to support multi-product producers with dynamic product configuration.
- Shipped bi-weekly production builds in an agile workflow, iterating on UI and backend based on feedback from 100+ live users.
Software Engineer Intern · Peoria Investments
Winter 2025
- Built Python automation scripts and reusable data pipelines to ingest and process property data, reducing manual evaluation time by 35% with unit tests ensuring data integrity.
Connect