AI Engineer: Generative, NLP, MLOps, Vision
Resume:
Introduction to Responsible AI
Introduction to Large Language Models
Introduction to Generative AI
Responsible AI for Developers: Interpretability & Transparency Responsible AI for Developers: Fairness & Bias
Machine Learning Operations (MLOps) for Generative AI Inspect Rich Documents with Gemini Multimodality and Multimodal RAG Vector Search and Embeddings
Introduction to Vertex AI Studio
Create Image Captioning Models
Transformer Models and BERT Model
Encoder-Decoder Architecture
Attention Mechanism
Introduction to Image Generation
14 Basics Certificates and Badges and 1 Intermediate Certificate about Machine Learning and Generative AI from GOOGLE:
CERTIFICATES
Developed custom AI chatbot solutions for small businesses to automate customer support and service handling.
Integrated AI systems to automate business processes such as data processing, internal communication, and task management.
Researched and implemented AI applications in financial trading using NLP and Large Language Models
(LLMs) for market analysis and news interpretation. AI Engineer
FREELANCE
Optimized model performance through evaluation, tuning, and integration into production systems. Built and deployed AI models for real-world use cases such as enterprise chatbots and educational assistants.
Developed teaching materials and demo projects on agentic AI for corporate training and internal adoption.
AI Engineer
WORKING CYBERSOFT TECHNOLOGY CO.,LTD
EXPERIENCES
University of Information and Technology
Bachelor in Computer Science
EDUCATION
AI Engineer with hands-on experience in NLP, Computer Vision, and Generative AI. Skilled in developing, fine-tuning, and deploying models in real-world applications. Participated in international AI competitions and contributed to practical AI solutions across multiple domains.
ABOUT ME
www.khaportfolio.info
Tan Binh District, Ho Chi Minh City
AI ENGINEER ******.****@*****.***
NGUYEN HUYNH HOANG KHA
LANGUAGES
Feb 2024 - Now
Nov 2022- Jan 2023
Considered as a great top 10 AI Developer in Google Cloud Skill Boost Github Linkedin
English ( IELTS 6.0 )
Vietnamese
SKILLS AI & Machine Learning Skills
Machine Learning & Deep Learning: TensorFlow,
PyTorch, Scikit-learn, XGBoost, LightGBM
Natural Language Processing (NLP): Hugging Face
Transformers, SpaCy, NLTK, BART, GPT, LLaMA
Computer Vision (CV): OpenCV, YOLOv8, Detectron2,
MMDetection
Model Optimization & Training: Hyperparameter Tuning
(Optuna, GridSearchCV), Bayesian Optimization
AI Deployment & MLOps
MLOps & Model Deployment: Docker, Kubernetes,
CI/CD, MLflow, FastAPI, Flask
Vector Search & Embeddings: FAISS, Pinecone,
ChromaDB, RAG Pipeline.
AI Serving & APIs: TensorFlow Serving, Triton Inference Server, RESTful APIs, GraphQL
Data Processing & Feature Engineering: Pandas,
NumPy, Dask, Polars
Mentor DA
AI instructor
Helping and imparting DA knowledges about Power BI and SQL related to DA June 2025 - Now
Helping and imparting DA knowledges about Machine Learning and Deep-Leanrning models about AI June 2025 - Now
MULTIMODAL VIDEO Q&A SYSTEM (RAG-BASED CHATBOT)
Model Researcher & AI Deployment
03/2024 - 08/2024
Designed and deployed a multimodal RAG-based Q&A system capable of processing video, text, image, and speech inputs for context-aware information retrieval. Integrated Whisper (speech-to-text), CLIP (vision-language), and Transformer-based embeddings for multimodal fusion.
Built an end-to-end pipeline: Inputs Preprocessing Embeddings (CLIP/Whisper/OpenAI) Vectorstores (FAISS/Chroma) LangChain + LangGraph orchestration Response generation Gradio interface.
Optimized retrieval accuracy (+40%) and latency (-30%), leading to improved user engagement in e- learning and knowledge management applications.
Delivered an interactive prototype with natural multimodal Q&A, showcased in the Advanced AI for Business program.
Tech Stack: Python, PyTorch, Whisper, CLIP, Transformers, FAISS, Chroma, OpenAI API, Google Colab, LangChain, LangGraph, GPT Models, RAG, Gradio, Embedding Models MULTI-HOP REASONING QA CHATBOT
AI Researcher
05/2025 - 08/2025
Built an MVP QA system with multi-hop reasoning, tracing reasoning chains over DBLP/Wikidata knowledge graphs.
Applied LLMs (LLaMA-3/Qwen-7B) for question decomposition (CoT) and GNNs (GraphSAGE/GAT) for multi-step relation encoding.
Orchestrated workflows with LangGraph, merging LLM + GNN outputs into final answers with step-by- step explanations.
Delivered an interactive prototype (FastAPI + Streamlit) with caching (SQLite/FAISS) for faster, auditable queries.
Tech Stack: LLaMA-3/Qwen-7B, PyTorch Geometric/DGL, LangGraph, DBLP/Wikidata, FastAPI, FAISS. MHQ-REACTRAG — RESEARCH-GRADE MULTI-HOP QA ASSISTANT Model Researcher & AI Deployment
03/2024 - 06/2024
Developed a research-grade multi-hop QA system combining Chain-of-Thought (CoT), ReAct loops, and RAG to provide verifiable answers with reasoning traces. Integrated tool-use capabilities (web search, vector retrievers, Neo4j KG, Python sandbox) and a FEVER-style verifier to ground claims and reduce hallucinations. Designed a science-grade pipeline with logging, benchmarks, ablation studies, and reproducible experiments for evaluation across QA datasets (HotpotQA, StrategyQA, GSM8K). Delivered an MVP with FastAPI backend + Streamlit/Gradio UI, supporting FAISS (dev) and scalable vector DBs (Milvus/Pinecone) in production.
Tech Stack: LLaMA-3/Qwen-7B, LangChain/LangGraph, PyTorch, FAISS/Milvus/Pinecone, Neo4j, Gradio. VIRTUAL - TRY ON 2D
Model Researcher & AI Deployment
03/2025 - 06/2025
Pipeline: User photo + garment image Human Parsing (segmentation) Pose Estimation U-Net image generation Image Warping Final try-on visualization. Built a 2D Virtual Try-On (VTON) system for e-commerce, allowing users to try clothes virtually using only a single image.
Leveraged U-Net for segmentation, OpenPose/MediaPipe for pose alignment, and image warping for realistic rendering.
Result: Produced realistic try-on outputs with high alignment quality; deployed as a demo for online fashion platforms.
Tech Stack: Python, OpenCV, PyTorch, U-Net, OpenPose/MediaPipe, Image Warping STUDENT’S BEHAVIOUR DETECTION
Model Researcher & AI Deployment
03/2025 - 06/2025
Pipeline: Classroom video YOLOv8 + MediaPipe for face/pose detection Behavior recognition
(drowsiness, yawning, distraction) Time-series analysis Prediction of academic performance. Designed a computer vision system to monitor student engagement in real-time. Combined behavioral cues (eye closure, yawning, head pose) with time-series prediction models to estimate attention levels and correlate with academic outcomes. Result: Generated early-warning alerts for at-risk students, helping educators improve classroom effectiveness.
Tech Stack: OpenCV, MediaPipe, YOLOv8, Python, Scikit-learn, CSV Time-series Analysis, Behavior Prediction Models
PERSONAL
PROJECTS
PUBLICATIONS /
PAPERS
Retrieval-Augmented Generation Architectures: Integrating LangChain, LangGraph, Vector Stores, and Open-Source Large Language Models -A Survey and Case Study with GPT-OSS-20B
Author: Kha Nguyen
Contact this candidate