Portfolio · 2026

I build AI systems,
and photograph
the world.

I'm interested in how AI systems plan, where they fail, and how to make them reliable — and a little more human.

BlindRun pairs blind runners with volunteer guides in real time. I built its backend alone; it is now in testing with community partners.

USACO Gold 2026
AIME qualifier 2025

Shenzhen Computer Science · Engineering GitHub

Jayden speaking at a developer conference, microphone in hand, a slide on data structure design behind him — Speaking at COSCon'25, the 10th China Open Source Conference, on my LLM task-planning research (SAER).

from the BlindRun case study

I run, and at a road race in Shenzhen I found blind runners doing the same. For many of them a guided run is one of the few standing reasons to leave the house, and the clubs that organize it had runners, had volunteers, and had no way to put the two together: coordination ran on phone calls and group chats, and volunteer certification ran on one organizer traveling city to city. I built the layer that was missing, then guided a run myself and rebuilt the parts that had quietly assumed a sighted user.

read the full case study →

BlindRun

The coordination layer that visually impaired runners and volunteer guides needed simply didn't exist — so I built it, backend first, then the app.

Java 17
Spring Boot
MySQL
Redis
WebSocket
Swift
SwiftUI

Relevance-Filtered Context Remediates the Task-Planning Bottleneck in Embodied LLM Agents

A retrieval method (SAER) that targets the real reason large language models collapse on multi-step tasks — and the ablation that showed the gain comes from somewhere other than where I thought.