Nguyen Phan Minh Triet

AI Engineer

AI Engineer specializing in agentic systems, MCP integrations, and LLM orchestration with domain experience in fintech and cybersecurity. Proven track record shipping multi-agent workflows and RAG pipelines at production scale.

View Work My Blog Resume
πŸ“§ [email protected] πŸ”— GitHub πŸ’Ό LinkedIn

Technical Skills

Languages Python C/C++
AI & LLM LangGraph Langfuse MCP Server LLM Integration OCR/Vision Prompting RAG Vector Embedding
Frameworks FastAPI Celery Pandas
Technologies MySQL Elasticsearch Redis Docker Kubernetes Linux Jupyter

Work Experience

AI Software Engineer

ZALOPAY - VNG Corporation Aug 2025 β€” Present
  • Architected & deployed a multi-agent ticket processing system via a LangGraph state machine. Processed 32K+ tickets with 80.1% automation rate peaking 99.8% for IBFT.
  • Designed 9+ smart workflows (IBFT, promotions, reopen handling) with LLM intent classification. Slashed handling time 4x faster, saving the CS team 187 hrs/mo.
  • Deployed and managed self-hosted LLM inference stack (Qwen 3.5, Gemma 3, PaddleOCR) on Kubernetes using NVIDIA NIM and vLLM, enabling cost-efficient, low-latency model serving for production agentic workflows.
  • Built a multi-model OCR/Vision platform achieving 99% accuracy, halved VietQR manual tickets (800 β†’ 400/mo) and cut CS workload from 66 to 33 hrs/mo.
  • Built MCP Registry, a centralized AI tool management platform. Currently in POC, scaling company-wide to empower engineers to contribute MCP servers/tools to the internal AI ecosystem.
  • Engineered a Code Generation Agent integrating an MCP Server & Figma-to-React workflow with Zalopay Looknfeel. Widely adopted by engineers for maximizing component reusability & reducing frontend maintenance.
LangGraph MCP Server vLLM NVIDIA NIM FastAPI Kubernetes PaddleOCR Python

Software Engineer

OPSWAT Jul 2024 β€” Aug 2025
  • Maintained and enhanced the security and performance of MetaDefender ICAP Server, an OSI layer 7 application implementing the ICAP protocol (RFC 3507) for network content adaptation and network security.
  • Designed and deployed multiple internal tools (IP Port Scanner, Jira Ticket Creator) on Kubernetes to automate workflows and reduce manual effort.
  • Implemented a monitoring system using Go, InfluxDB, and Grafana to track CPU, memory, disk, and network usage across multiple processes, helping detect memory leaks, identify CPU anomalies, and compare performance across releases.
  • Supported global sales engineers in deploying ICAP server for 10+ enterprise clients, integrating with data loss prevention tools and web proxies.
  • Improved HTTP/ICAP multi-threaded parsing, increasing throughput by 25% and reducing response time by 30% via refactoring chunked request logic and minimizing resource contention.
  • Led PostgreSQL migration from v12.2 to v16.6 across Linux and Windows, resolving compatibility issues and ensuring zero data loss.
C/C++ Python Go ICAP PostgreSQL InfluxDB Grafana Kubernetes Linux

Research Collaborator | Part-time

Young Monkeys May 2024 β€” Jul 2024
  • Contributed to the open-source organization Young Monkeys in Vietnam through the EzyPlatform project, enabling users to build software with minimal programming by integrating plugins.
  • Received recognition for significant contributions on the contributors' page and in a heartfelt dedication, lauding diligent research and documentation efforts that amplified the accessibility and utility of EzyPlatform.
Java EzyPlatform Open Source

Software Engineer Internship

Zodinet Technology May 2024 β€” Jul 2024
  • Gained hands-on experience with full-stack development using Node.js, ReactJS, and NestJs.
  • Received structured training in software development workflows, collaboration tools, and client communication.
Node.js ReactJS NestJs Full Stack

Research Collaborator | Part-time

Computer Vision Project - University of Oxford Apr 2023 β€” Jul 2023
  • Collaborated with a PhD student under Prof. Stephen Baker's guidance, engaging in dataset labeling and literature reviews.
  • Actively participated in project meetings, presented research findings, and provided weekly progress reports to the mentor.
Computer Vision Research Dataset Labeling

Backend Collaborator | Part-time

DATAFIRST Sep 2022 β€” Nov 2022
  • Developed a real estate pricing estimation project, enhancing accuracy by 60% and overcoming rate-limiting on Cho.com.
  • Utilized Python Selenium for data scraping from Cho.com and Bat.com, and implemented data extraction from Facebook groups using scrapy-splash.
Python Selenium scrapy-splash Data Scraping

Projects

Fullstack, Book Store Management System

Jan 2024 β€” Apr 2024

Team Size: 4

Tech Stack: Java Core, JavaFx, JDBC, MySQL

Key Achievements:

  • Modular design and effective user role management.
  • Implemented CRUD operations with streamlined controllers.

Challenges and Solutions:

  • Segregated user roles into distinct packages for clear management.
  • Employed utility classes for efficient input validation and CRUD operations.
GitHub

Developer, Fashion Revive

Jul 2023 β€” Sep 2023

Team Size: 4

Technologies: ReactJS, NodeJS, MongoDB, S3 Bucket, Vnpay, React-oauth, Tailwind CSS

Key Achievements:

  • Integrated user authentication, fortifying platform security and increasing user trust.
  • Implemented separate dashboards for buyers and sellers, enhancing user experience.
  • Efficiently listed, searched and filtered products, improving discoverability by 25%.

Challenges and Solutions:

  • Encountered the challenge of integrating third-party APIs to ensure secure and efficient transactions; resolved by efficiently utilizing the VnPay API.
  • Ensured optimal performance by fine-tuning components and optimizing backend queries.
GitHub Demo

Education

Information Technology, HCMUS-UCBL1 Double Degree Program

VNUHCM – UNIVERSITY OF SCIENCE Sep 2021 β€” Nov 2025

Relevant Coursework:

  • DSA
  • Software Engineering
  • Database Systems
  • Web Development
  • Computer Networks

Achievements & Certifications

πŸ†

Best Employee of Q2 2025 – OPSWAT

Recognized as Best Employee for outstanding contributions to product performance, security, and client deployments.

πŸ₯‡

First Prize – Zalopay CTF 2025

Achieved first place at the Zalopay Capture The Flag 2025 competition.

πŸ“Š

Big-O Blue Intermediate Algorithms

Completed advanced algorithms and data structures certification program.

View Certificate
πŸ€–

IBM RAG and Agentic AI Professional Certificates

Completed IBM's professional certification programs in RAG and Agentic AI systems.

View Certificate
⭐

Open Source Contributor

Active contributor to the Young Monkeys open-source organization in Vietnam through the EzyPlatform project.

View Contributions
🎯

Project Excellence

Recognized for outstanding project delivery and technical innovation.

View Post

Get In Touch

I'm always interested in new opportunities and exciting projects. Feel free to reach out if you'd like to collaborate or just have a chat!

πŸ“§
Email

[email protected]

 πŸ’Ό
LinkedIn

Connect with me