Ai Engineer Product Owner
Resume:
Hoàng Ngọc Tiến
AI Engineer H T
LICENSES &
CERTIFICATIONS
TECHNICAL SKILLS
EDUCATION
FSOFT Stars 2024
• FSOFT Stars 2024 Award for
employees with outstanding
achievements
NVIDIA
• NVIDIA Certified Associate:
Generative AI LLMs
• Building RAG Agents with LLMs
Building Real-Time Video AI
Applications
• Introduction to Transformer-Based
Natural Language Processing
NVIDIA Getting Started with AI on
Jetson Nano
IIG Vietnam
• TOEIC – 800
Ho Chi Minh City University
of Technology
Major: Aerospace Engineering
GPA: 8.19/10
September 2016 – 2020
Major: Computer Science (Master)
January 2023 – Now
OBJECTIVE
Short-term:
Become a manager to lead product
development and deliver value-
driven solutions.
Long-term:
Create a best-selling
technological product for the
global market.
EXPERIENCE
FPT Software Oct 2023 – Now
Develop a comprehensive AI Platform named IvyChat serving diverse industries including BFSI and manufacturing.
July 25, 1998
Male
District10, Hochiminh, Vietnam
****************@*****.***
linkedin.com/in/hoang-ngoc-tien/
Language: Python
Frameworks: Tensorflow, Pytorch
DBMS: Mongodb, MySQL, Postgres
Others: VLLM, ONNX, Triton Server,
Docker, Jenkins, Langgraph, Langchain
Develop a chat bot that integrates with SQL databases, allowing users to interact naturally by asking questions in their preferred language.
Description: This platform integrates multiple advanced modules tailored for enterprise needs: Intelligent document interaction, Text-to- SQL and Visualization, Multimodal AI Generation, Multi-agent system. The product brought in $5M in revenue in the first year.
Responsibilities:
- Managing and consulting engineering team.
- Communicating with stakeholders to align on product requirements and objectives.
- Design multi-agent systems using LangGraph for complex enterprise solutions, including product recommendation and risk assessment management.
- Develop Intelligence Document Interaction with LangChain and Azure/AWS, tailored for diverse industry needs.
- Develop Real-time Voice Assistant with multimodal input and output.
- Fine-tuning open-source models (e.g., Gemma2 7b, Llama 3.1 8b) to replace cloud-based solutions.
- Develop Deep-Research features by utilizing reasoning models (e.g., deepseek-r1, o3-mini) tailored for specialized research tasks.
- Build on-premise systems integrated with NVIDIA’s tech stack (NIM, NEMO, NV-Ingest) to leverage NVIDIA hardware capabilities. Skills: Product Owner · Team Leadership · Customer Communication · Multi-agent Systems · Large Language Models (LLM) · RAG · NVIDIA
Description: Empower users to interact with enterprise SQL databases using natural language, leveraging domain-specific knowledge to generate accurate queries without requiring SQL expertise. The first product at AI center sold as a license model.
Responsibilities:
- Managing and consulting engineering team
- Developing and executing project timelines to ensure on-time delivery. EXPERIENCE
TMT Solutions Aug 2021 – Sep 2023
Build services relevant to image processing for fashion store sales consulting Chatbot.
Description: Search for similar products in the database from an image sent by user. It’s a function of TMTChatbot.
Responsibilities:
- Responsible for coding whole service. Working with the Product Owner to define the necessary functions.
- Working with the chatbot team to code the required APIs.
- Custom and training image embedding models (TokenMix), object detection models (Yolov8), and classification models.
- Convert the model to ONNX and deploy on Triton server that increase 50% performance. Skills: TokenMix · Yolov8 · Vision-Transformer · Qdrant · FastAPI · Kafka · Docker Develop an Optical Character Recognition (OCR) system.
Description: Using for extract information from house ownership certificate and supply for real estate company, extract information from various types of bills and supplies for shipping companies.
Responsibilities:
- Leader of the group of 4 members. Working with stakeholder to understand the requirements, define the functions, plan for the project and team.
- Project’s architecture design. Evaluates the results of each member, review code to ensure code quality.
- Perform data processing, train model VietOCR and Text Detection, evaluate model performance and accuracy. Skills: VietOCR · PaddleDetection · SDMGR · FastAPI · Docker · MongoDB Virtual try on clothing.
Description: Serving for customers who want to try on clothes, the model will generate an image of person wearing that product.
Responsibilities:
- Leader of the group of 3 members. Working with Product Owner to understand the requirements, define the functions, plan for the project and team.
- Working with data team to collect data for training.
- Train model stable diffusion with task inpainting, evaluate result and build Docker for this service. Skills: Stable Diffusion · FastAPI · Kafka · Docker Uniduc JSC Dec 2019 – July 2021
Face Recognition System.
Description: A timekeeping system use facial recognition technology. It have able to recognize multiple people in the same time.
Responsibilities:
- Working with human resources to synchronize data and use result of face recognition system for timekeeping.
- Define architecture, document.
- Write module get frame from multiple camera.
- Augment data to fit with real data and train model Partial FC. Skills: Partial FC · Retinaface · RabbitMQ · Docker · MongoDB
- Design SQL agent to incorporate additional client database information and business rules for enhanced quality.
- Evaluate the agent using customer datasets to ensure optimal performance. Skills: Product Owner · Team Leadership · Customer Communication · Langgraph & Lanchain · RAG
Contact this candidate