Carlos Castillo

Software engineer and quantitative trader building reliable systems at the intersection of financial markets, data engineering, and machine learning. I turn complex data into scalable software, rigorous analysis, and actionable decisions.

Projects

Synthetix Trading Engine

High-performance, event-driven C++20 engine featuring a hardware-optimized order book, real-time microstructure signals, and an institutional risk gatekeeper.

Gym Engine

A full-stack gym management platform built with a decoupled NestJS and Next.js 16 architecture, designed for secure access, operational reliability, and scalable member management. The system includes JWT-based role-based access control, a first-access onboarding flow for streamlined password resets, time-gated workout visibility through server-side logic, and a real-time attendance swap protocol backed by PostgreSQL and Prisma. I also built a custom HTML5 Canvas image compression utility to reduce storage costs and keep the dashboard fast and responsive.

Earnings Call
Analyzer

An automated NLP pipeline designed to turn earnings call transcripts into structured market insights. It extracts sentiment, key financial signals, and event-driven reactions before generating research-ready PDF reports, dramatically reducing manual analysis time.

Blog Application

A full-stack blog platform built with secure JWT authentication, GraphQL and REST APIs, SSR/SEO enhancements, analytics tracking, and real-time WebSocket room messaging. I engineered the backend services, led the Apollo migration, and containerized the application with Docker for deployment on Google Cloud Run, delivering a scalable and interactive publishing experience.