Who I am
I'm Sia Khorsand (he/him), a software engineer based in Irvine, California. I studied Cognitive Science and Computer Science at UC San Diego. I like working across a whole product, from deciding what to build to shipping it and supporting the people who use it.
Most of my work sits where AI systems meet real workflows: structured LLM outputs, agentic pipelines, backend APIs, and native iOS apps. The cognitive science background shows up in how I design interfaces and in how much time I spend watching people actually use what I build.
What I have built
Chalky (February 2026 to present)
I founded and built Chalky, a gym log and form-check app that is live on the App Store. It turns free-form workout notes into structured data, insights, and coaching, and reviews lift form from video. I own it end to end: research, product, design, the native SwiftUI app, the Supabase backend, the offline-capable sync layer, launch, and user support. Before launch I ran a TestFlight beta with more than 60 testers and reworked onboarding and failure states based on where they got stuck. The Chalky case study goes through the decisions in detail.
Uniwise (May 2025 to February 2026)
I was one of the first three engineers at Uniwise, building AI tools for organ-donation teams. I took two products from technician interviews to deployed MVPs: a live call copilot that streams labeled transcripts over WebSockets, and ContraSight, which reads scanned donor records with OCR and checks them against 71 contraindications, citing the exact source text for every finding. I worked across the full stack, mostly backend and AI (Python, FastAPI, MongoDB, S3, Auth0, Gemini), plus React and TypeScript front ends and AWS deployment. You can try an interactive recreation of both products built on fictional data.
UC San Diego (September 2024 to June 2025)
I was an instructional assistant for two Cognitive Science courses, COGS 150 (Large Language Models and Cognitive Science) and COGS 10 (Cognitive Consequences of Technology), supporting more than 100 students through labs, discussion sections, office hours, and project feedback.
How I work
I start by talking to the people who will use the thing and mapping how their work actually runs, then scope the smallest version that proves the idea. On the engineering side I care about failure modes that stay recoverable: structured outputs with validation, provider failover, idempotent writes, caching, and server-side quotas. I prefer policy code to own scoring and safeguards while models handle interpretation and drafting.
The portfolio homepage lists projects, skills, and experience. The contact page has the best ways to reach me, and the resume is a one-page PDF.