Michael – AWS, Python, React

A reinforcement-learning agent running as an autoscaled bot fleet to achieve superhuman slither.io play.

• Built a reinforcement-learning agent using Ray RLlib and PPO.
• Ran the agent as an autoscaled bot fleet to achieve superhuman Slither.io play.

An MMO server for slither.io with integrated crypto and fiat payment rails, enabling real‑money gameplay and scalable multiplayer for thousands of concurrent players.

• Reverse-engineered the Slither.io binary WebSocket protocol in 6 weeks to support custom MMO infrastructure.
• Designed and implemented a custom MMO server to handle real‑time Slither.io gameplay at scale.
• Integrated crypto and fiat payment rails to support deposits, withdrawals, and player account balances with real‑money stakes.

Enterprise SaaS platform for AI‑powered inventory automation and digital‑twin supply chains, helping warehouse and logistics teams replace manual stock checks with real‑time, high‑accuracy inventory insights

• Architected and built the end-to-end computer-vision pipeline processing hundreds of terabytes of 5K drone video.
• Built the full backend and DevOps stack on AWS using EC2, Fargate, S3, and RDS.
• Implemented the platform with Docker, MariaDB, an internal API, and GitHub CI/CD.
• Authored the core internal Python libraries for the data layer and database/ORM tooling used across the pipeline.
• Turned raw drone imagery into structured, reconciled warehouse inventory counts at enterprise scale for Fortune 500 customers.

Enterprise SaaS platform for AI-powered inventory automation, deployed across Southeast Asia.

• Architected and built the end-to-end CV pipeline processing 100s of TB of 5K drone video (YOLOv8, PyTorch, OpenCV, CUDA, Python).
• Built the full backend and DevOps stack: AWS (EC2, Fargate, S3, RDS), Docker, MariaDB, internal REST API, and GitHub CI/CD.
• Authored the core internal Python libraries for the data layer and database/ORM tooling used across the pipeline.
• Raised > $1m USD in government grants and VC investment.
• Turned raw drone imagery into structured, reconciled warehouse inventory counts at enterprise scale for Fortune 500 customers.

A GIPS-compliant performance measurement and compliance system for a $30B asset manager.

• Served as lead developer on the compliance system for the Global Investment Performance Standards.
• Wrote SQL for the compliance system.
• Linked reporting through Talend ETL and Excel VBA, Access, Crystal Reports, and SSRS.
• Architected the firm’s updated performance analysis system.
• Reported on performance issues and trends in bond, equity, and currency markets.