012022 – 2025
revu.com
VTWO Software — Senior Engineer
Problem: The platform pays users and affiliates for completing offers and surveys. It was being exposed to fraudulent traffic and to slow admin tooling. Fraud review had no tiered response and no audit trail. Concurrent first-time user creation caused database errors. The admin report for flagged users became slow to load as data grew.
Outcome: Built a three-tier fraud-control system (Good / Suspicious / Ghost-ban) on Verisoul. It diverts suspicious clicks, suspends credits, automatically escalates to a full ban, and logs every status change. If the vendor goes down, an uptime checker reverses any actions taken during the outage. Added PostHog feature flags so features ship safely without redeploying. Fixed first-time user creation errors under concurrent requests, which removed a source of 500 errors. Added pagination, sorting and an index to the flagged-users report, so it stays fast as data grows.
- PHP
- Laravel
- WordPress
- React
- GCP
- Docker
- Kubernetes
022022 – 2025
Cars.com
Innova Solutions — Tech Lead
Problem: Cars.com's PHP platform needed to scale reliably across cloud infrastructure while a cross-functional team kept shipping new features on tight sprint cycles.
Outcome: Directed containerization and orchestration on GCP with Docker and Kubernetes for high availability, while running SCRUM delivery and mentoring the team on architecture and backend practices.
- PHP
- Laravel
- WordPress
- React
- GCP
- Docker
- Kubernetes
032019 – 2022
Order My Gear
Lumina Software — Tech Lead
Problem: Order My Gear needed its e-commerce stack (WordPress and PrestaShop) to interoperate cleanly with custom back-end services, without slowing down feature delivery.
Outcome: Built Go microservices to bridge WordPress and PrestaShop, deployed scalable GCP infrastructure with Docker/Kubernetes, and designed CI pipelines that improved code quality and deployment speed.
- Go
- Laravel
- PrestaShop
- VueJS
- ReactJS
- GCP
- Jenkins
042016 – 2019
Kavak AI / CEDIS
Savant AI — Director of IT
Problem: Vehicle entry/exit verification at the company's distribution centers (CEDIS) was a manual process that didn't scale with volume.
Outcome: Led 9 engineers building an AI-driven automation system combining IoT devices, computer vision, and machine learning, backed by AWS infrastructure and real-time data pipelines — cutting manual verification time by over 60%.
- Golang
- Python
- PHP
- TensorFlow
- AWS
- ESP32
