Luan – Laravel, RabbitMQ, AWS

A licensed securities exchange needed one reliable way for partner brokers and investor apps to list instruments, stream live market data and place trades. Luan owned this partner integration layer end to end. He designed and built the partner trading API in Next.js, TypeScript and Prisma (with Swagger docs), covering accounts, balances, instruments, orders (market, limit and MFK), trades, transactions and the order book, with self-trade checks and idempotent execution-report webhooks. He also built a TypeScript FIX client that connects to the exchange's order and market-data sessions and publishes events to Redis Streams with RabbitMQ as a fallback, plus a stream-consumer service that delivers them to partner webhooks with automatic failover, exponential-backoff retries and per-partner routing. As a result, partners connect to the exchange through a single documented integration and get trading events reliably even when a component fails.

• Designed the architecture of the exchange's partner integration layer
• Built the core partner REST API in Next.js, TypeScript and Prisma, documented with Swagger
• Built and ran multiple FIX clients for the exchange's order and market-data sessions
• Built fault-tolerant, high-throughput event streaming with Redis Streams and RabbitMQ
• Delivered trading events to partners via idempotent webhooks with failover, retries and per-partner routing
• Containerized all services with Docker and shipped them through GitLab CI

A workforce-management SaaS for labor contractors that staff trade shows and live events. Coordinators plan labor per job and booth, send availability requests and dispatch union crews, and on-site time tracking feeds time sheets, payroll deductions and billing based on contract rules such as union rules, holidays and overtime. As Lead Software Architect and the second-largest contributor to the codebase, Luan led a team of 4 developers and worked across a Laravel API, an Angular web app and a mobile API on AWS. He built QR-based on-site check-in/out with geofencing and time sheets, labor planning and dispatch with contract rules, and rebuilt email, SMS and push notifications on AWS Pinpoint with a self-serve notification-rules system. He also moved login from a legacy SSO server to AWS Cognito, moved the mobile API to ECS, added Datadog APM monitoring and redesigned the CI/CD release flow so teams can deploy on their own.

• Led a team of 4 developers across front-end, back-end and mobile
• Built QR-based on-site check-in/out with geofencing, time tracking and time sheets
• Built labor planning, availability requests and crew dispatch based on contract rules
• Rebuilt email, SMS and push notifications on AWS Pinpoint and added self-serve notification rules
• Migrated authentication from a legacy SSO server to AWS Cognito
• Moved the mobile API to Amazon ECS and added Datadog APM monitoring
• Redesigned the CI/CD release flow and maintained infrastructure with Terragrunt, AWS and GitHub Actions

A software development company delivering web, mobile and machine-learning projects for a range of clients. Luan worked across the stack, building Angular, React, Ionic and Vue.js front ends and Laravel and Node.js back ends, and led the design and architecture of projects and features while mentoring other developers.

• Built web and mobile front ends with Angular, React, Ionic and Vue.js
• Built back ends and APIs with Laravel and Node.js
• Worked on machine learning and deep learning projects
• Led the design and architecture of projects and features
• Mentored developers and reviewed code
• Ran code releases and production deployments

A research and technology institute's project to classify footprints as pronated, supinated or neutral, so the results could be used to produce custom-made insoles. Luan worked as a data scientist and machine learning engineer and built the deep learning model behind the classification.

• Built a deep learning model that classifies footprints as pronated, supinated or neutral
• Worked as a data scientist and machine learning engineer on the footprint data

Research projects at a technology institute applying AI to real-world problems. The main project used deep learning and optimization algorithms (Particle Swarm Optimization, Genetic Algorithms) to solve the seismic full waveform inversion problem in geophysics, which helps locate petroleum and other materials under the subsurface. Luan also worked on recurrent neural networks that predict the outcome of legal cases.

• Applied Particle Swarm Optimization, Genetic Algorithms and deep learning to seismic full waveform inversion
• Built GRU and LSTM recurrent neural networks to predict the outcome of legal cases

A university research project to build a colony of robots that navigate an environment on their own, explore it, and identify and interact with other robots. Luan worked on the navigation part, using fuzzy logic systems in a simulated environment.

• Built fuzzy logic systems for mobile robot navigation in a simulated environment