Brian - MLOps Engineer
👋

Hi, I'm Brian

I design and build production-grade ML systems — from automated training pipelines to real-time inference infrastructure. Bridging the gap between data science and reliable software engineering.

Scroll to explore

About me

I am an MLOps Engineer with over 6 years of hands-on experience, blending expertise in AI model research, backend development, and application deployment. My passion lies in leveraging cutting-edge technologies to create impactful solutions.

Known for my adaptability and quick learning, I thrive in dynamic environments and embrace new challenges. I seek opportunities to design and deploy innovative AI solutions that address real-world challenges, from OCR pipelines to intelligent camera systems.

AI Model Research

Backend Development

Production Deployment

OCR & Computer Vision

Skills

Technologies and tools I work with

Container & Orchestration

  • Kubernetes
  • Docker
  • Helm
  • Argo Workflows

ML Frameworks

  • PyTorch
  • TensorFlow
  • Hugging Face
  • scikit-learn

MLOps Tools

  • MLflow
  • Kubeflow
  • Airflow
  • Prefect

Cloud & Monitoring

  • AWS
  • GCP
  • Prometheus
  • Grafana

Languages & Frameworks

  • Python
  • TypeScript
  • FastAPI
  • Next.js

Databases & Messaging

  • PostgreSQL
  • Redis
  • Apache Kafka

Experience

  • Senior MLOps Engineer

    CMC Global

    ·Hanoi, Vietnam

    Building a full-cycle MLOps OCR pipeline for the internal team. Lead OCR projects for multiple clients (Australia markets) — solution adapts to a variety of templates with 2 seconds/page performance. Integrate and deploy a customer scoring model in a self-hosted Kubernetes cluster for a banking client. Manage team of 3–6 members, design solution architecture, and contribute hands-on to critical modules.

    PythonPyTorchFastAPIvLLMComputer VisionBash ScriptDockerKubernetes
  • AI Team Lead / MLOps Engineer

    Viettel High Tech

    ·Hanoi, Vietnam

    Built an AI serving system for cameras handling RTSP/HLS streams and third-party smart home events. Built systems to manage model and solution versions for the serving system and edge devices. Reduced resource consumption by ~30% while maintaining processing performance. Deployed on-premise Kubernetes cluster with high availability and security. Promoted from MLOps Engineer to AI Team Lead in October 2023.

    PythonPyTorchONNXFastAPICeleryTriton Inference ServerDeepStreamMediaMTXRedisKafkaRabbitMQMongoDBMilvusDockerKubernetesHelm
  • MLOps/AI Engineer (Part-time)

    MachinaX

    ·Hanoi, Vietnam

    Built AI solutions meeting user requirements across projects including Hair Style Manipulation, OCR for drilling pipes, and Virtual Speaking Room modules. Designed full pipeline solutions, trained AI models, packaged and deployed to customer environments on-premise and on AWS. Created test datasets and reported results to clients.

    PythonPyTorchFlaskCeleryRay ServeBash ScriptDockerAWS
  • Backend Engineer

    CETI

    ·Hanoi, Vietnam

    Main contributor to the backend of an ERP platform for an oil corporation. Developed a desktop application integrating real-time data via MQTT and Firebase.

    Java Spring BootJavaFXSQL ServerMQTTFirebase
  • AI Engineer

    Personal Project

    ·Hanoi, Vietnam

    Developed an OCR solution for a banking company that extracts information in predefined categories from documents such as ID cards and Passports. Researched AI models, enhanced results with pre/post-processing algorithms, and created a test dataset.

    PythonPyTorchOpenCVDeep LearningDocker
  • Computer Vision Engineer

    Vietnam Artificial Intelligence System

    ·Hanoi, Vietnam

    Researched and built a solution to remove objects from studio background images. Prepared data (collect, clean, analyse), experimented with fine-tuning strategies and pre/post-processing to improve overall performance, and reported results to the team leader.

    PythonPyTorchOpenCVDeep LearningBash Script

Contact

Open to new opportunities, collaborations, or just a tech discussion. Reach out directlyand I'll get back to you.

Click to copy