Modelling & evaluation
AI / Machine Learning
Training, fine-tuning and evaluating models — from classical baselines to frontier architectures.
- Python
- PyTorch
- scikit-learn
- TensorFlow
- Keras
- NumPy
- OpenCV
- Hugging Face
- Onnx
- Google Colab
AI/ML Engineer building agents, computer vision that ship to production with scalable on cloud.
Currently
Open to new work
I bridge the gap between experimental AI and resilient, production-grade cloud systems.
As an AI/ML Engineer specializing in autonomous agent workflows and computer vision, my focus is not just training models, but deploying them where they matter—at scale, with low latency, and zero downtime. I design end-to-end architectures that take research off the notebook and turn it into robust, cloud-native products capable of serving real-world users reliably.
Grouped by discipline — each tile is a domain I work in day to day, with the tools I reach for inside it.
44 tools · 7 domains
Modelling & evaluation
Training, fine-tuning and evaluating models — from classical baselines to frontier architectures.
Agent systems
Chunking, embeddings, hybrid search, tool callings and the eval harnesses that keep answers honest.
Serving layer
Typed, tested services that put a model behind a stable contract.
database
Ingestion, transformation and the stores that keep features and vectors queryable.
Ship and observe
Containers, CI/CD, experiment tracking and the monitoring that catches drift early.
automation & pipeline
Building automation workflows, background process, gathering data in to warehouse
Collaboration
Planning features, tasks tracking, issue reports, working with teammate
Each card opens the full story — what the project set out to do, what I owned, the numbers it moved, and the stack behind it.
012025 — 2026AI Engineer
An interactive geospatial AI copilot enabling users to query, analyze, and manipulate geographic layers and spatial points via natural language.
022025 — 2026Data Engineer
An automated scheduling and routing optimization pipeline feeding enterprise field data into Timefold to assign and visualize technician dispatches.
032025 — presentAI Engineer & Fullstack Engineer
A multimodal medical assistant supporting clinical data retrieval, dietary nutrition extraction from images, and daily activity tracking from patient audio streams.
042025 — presentFounder, AI Engineer, Fullstack & Data Engineer
A legal-tech agent platform tailored for Vietnamese boutique law firms, automating legal research over statutory frameworks and drafting procedural documents.
Formal qualifications and the certifications behind the tools I work with — each one links out to the issuer's copy.
2 credentials · 1 degree
FPT University — Da Nang
Artificial Intelligence · GPA 3.5 / 4.0
Graduation thesis on time-series imputation on a Vietnamese missing dataset.

Claude Academy
Open to AI/ML engineering roles, research collaborations, and freelance builds. The fastest way to reach me is email — I reply within a day or two.