Backend Engineer · AI Researcher · Remote
Open to remote backend / ML roles — hxtruong6@gmail.com
- 🔬 osr-metrics — Open-Set Recognition & OOD-detection metrics for trustworthy ML.
- 📚 Reading deeply on Bayes-optimal multi-label prediction and pre-order label structures.
Backend engineer by trade, AI researcher by interest.
Day job — Backend. Production services with NestJS + TypeScript, PostgreSQL, and object storage on MinIO/S3. Comfortable across REST/GraphQL APIs, auth, queues, caching, observability, containerization, and CI/CD.
Research — Trustworthy & uncertainty-aware ML:
- 🔬 Open-Set Recognition (OSR) & OOD detection — how models behave on inputs they were never trained for
- 🏷️ Multi-label classification — Bayes-optimal prediction, evaluation metrics, pre-order structures over labels
- 🕸️ Graph embedding — representation learning on dynamic graphs
- 👁️ Applied computer vision — detection pipelines (YOLO family), image classification, fault detection
Backend (primary)
Infra & DevOps
Research / ML
🔬 Research
| Project | What it is |
|---|---|
| osr-metrics | Open-Set Recognition & OOD-detection metrics for ML research |
| inference_probabilistic_mlc | Bayes-optimal-prediction inference for multi-label classification |
| pre-order-for-mlc | Pre-order structures over labels for multi-label tasks |
| net2net | Network widening/deepening for incremental capacity |
| dynamic-graph-embedding | Embedding methods for dynamic graphs |
👁️ Applied CV
| Project | What it is |
|---|---|
| fe-credit-card-fault-detection | Card ID fault detection (7★) |
| butterfly-yolov3 · brand-detection | YOLOv3 detection pipelines |
🛠️ Backend & systems
| Project | What it is |
|---|---|
| go-ecommerce-microservices | E-commerce microservices in Go |