Fullstack Developer with GenAI & Computer Vision

SUMMARY

TECHNICAL SKILLS

Programming Languages: Python, JavaScript, TypeScript, C/C++

Frontend: HTML5, CSS3, React.js, Next.js, Tailwind CSS, Ant Design Backend & APIs: Node.js, Express.js, NestJS, RESTful APIs Databases & Data: MySQL, PostgreSQL, MongoDB, Prisma, Firebase Realtime Database AI / Machine Learning: PyTorch, Hugging Face Transformers, OpenCV, Edge Impulse, Natural Language Processing, Computer Vision

LLM & Generative AI: OpenAI API, LangChain, Prompt Engineering, LLM Application Development, Structured Outputs

Automation & Integration: n8n, Workflow Automation, API Integration, Notifications, Logging, Scheduled Workflows, Analytics Reporting

Computer Vision: Image Classification, Object Detection, Image Segmentation, Feature Extraction, Image Retrieval

Developer Tools: Git, GitHub, Postman, Docker

Operating Systems: Windows, macOS, Linux (Ubuntu)

WORKING EXPERIENCE

XHEROZONE Fullstack Developer March 2026 - June 2026 Project: CRM & Admin Workflow Management

Description: A comprehensive web application integrating Admin and CRM modules to streamline consultation workflows across Sales, Branch Operations, Survey, Design, QC, and Purchasing. Responsibilities:

Developed and optimized CRM and Admin workflow features to route consultation profiles across Sales, Branch Operations, Survey, Design, QC, and Purchasing, while implementing manual and automated task scheduling for the Design team based on office hours and Team Leader approval workflows.

Built a multi-level design submission and acceptance process involving Design Staff, Team Leader, Department Head, and QC, ensuring outputs followed the required approval flow before finalization and hard-copy printing.

Developed a real-time Livestream Management module supporting session creation, configuration, stream list management, and consultation forms, and integrated Firebase Realtime Database to provide low-latency synchronization across livestream-related features. Developed frontend interfaces using React.js, Tailwind CSS, and Ant Design, and integrated them with Node.js services, MongoDB, and Firebase Realtime Database for application data handling and real-time functionality.

Technologies: React.js, Tailwind CSS, Ant Design, Node.js, MongoDB, Firebase Realtime Database. Ho Chi Minh City • 089******* • ****************@*****.*** • Github • Linkedin NGUYỄN HỒNG QUẾ ANH

SOFTWARE ENGINEER FULL-STACK • APPLIED AI • AUTOMATION Software Engineer with hands-on experience building full-stack web applications, workflow-driven systems, real-time features, automation pipelines, and applied AI solutions across web development, NLP, computer vision, and embedded AI.

Experienced with React.js, Next.js, TypeScript, Node.js, NestJS, RESTful APIs, relational and NoSQL databases, as well as AI/ML technologies including Python, PyTorch, Hugging Face Transformers, OpenAI API, LangChain, and OpenCV.

Strong ability to translate operational, technical, and creative requirements into maintainable applications and reliable workflows involving real-time synchronization, structured outputs, API integrations, notifications, logging, approval processes, and analytics reporting. PROJECTS

E-Commerce Admin Dashboard & Store (Currently In Development) Fullstack Web Technologies: Next.js, TypeScript, Node.js / NestJS, PostgreSQL, Prisma, Tailwind CSS Developed a full-stack e-commerce platform consisting of a customer-facing storefront and administrative dashboard, with reusable and responsive UI components built using Next.js, TypeScript, and Tailwind CSS.

Implemented authentication, role-based access control, and product, category, inventory, and order management modules while integrating frontend functionality with RESTful APIs and backend services.

Designed relational PostgreSQL schemas using Prisma ORM and implemented server-side data fetching to support application data and administrative workflows. AI Creative Asset Workflow Automation AI Automation Technologies: Workflow Automation, Prompt Engineering, LLM APIs, Cloud Storage, Email, Slack Built an end-to-end workflow that reads structured descriptions, reference asset URLs, output formats, and model specifications, then automates asset generation and organizes generated outputs in cloud storage.

Implemented success and failure notifications through email and Slack, execution status and error logging, and daily reporting with success-rate and error-rate analytics for workflow monitoring.

Designed and iterated prompts for asset generation by comparing generated outputs against references, documenting rationale, and refining prompt instructions based on observed results. AI-Based Fruit Recognition using Edge Impulse on ESP32 AIoT & Computer Vision Technologies: Python, Edge Impulse, ESP32-CAM, OpenCV, ESP Mail Client, SPIFFS Developed an AI-powered fruit recognition system on ESP32-CAM using Edge Impulse for on- device image inference, including image capture, AI inference, local SPIFFS storage, and deployment of a trained classification model.

Implemented Wi-Fi connectivity and SMTP email services to send captured images when fruit was not detected, integrating the ESP32 Camera library, Edge Impulse SDK, and ESP Mail Client for real-time embedded computer vision.

AI VoiceBot for the Visually Impaired NLP

Technologies: Python, OpenAI API, LangChain, ElevenLabs, Speech Recognition Developed an AI-powered voice assistant combining speech-to-text, natural language understanding, LLM-based question answering, conversational memory, and text-to-speech to provide conversational assistance for visually impaired users. Designed voice-driven navigation and hands-free interaction, integrating speech recognition, OpenAI API, LangChain, and ElevenLabs into a complete conversational pipeline. Named Entity Recognition using Transformer Models NLP Technologies: Python, PyTorch, Hugging Face Transformers Fine-tuned transformer-based language models for Named Entity Recognition on the CoNLL- 2003 dataset and developed preprocessing and tokenization pipelines for sequence labeling. Evaluated model performance using Precision, Recall, and F1-score to measure entity recognition quality.

Image Retrieval using Deep Hashing (CIFAR) Computer Vision Technologies: Python, PyTorch

Implemented a deep hashing model for large-scale image retrieval, learning compact binary feature representations for efficient nearest-neighbor search and applying deep learning techniques for feature extraction and similarity matching. Evaluated retrieval performance using Mean Average Precision (mAP) to assess image retrieval quality.

LANGUAGES

English: Upper-Intermediate (CEFR B2, VSTEP 7.5)

Certification: VSTEP certification 7.5-B2 [ref] [view] Chinese: Beginner (HSK 1)

EDUCATION

Aug 2021 - Dec 2025

Bachelor of Science in Computer Science (Specialization in Natural Language Processing) Relevant Coursework: Data Structures and Algorithms Object-Oriented Programming Database Systems Operating Systems Artificial Intelligence Machine Learning Natural Language Processing Computer Vision

University of Science, Ho Chi Minh National University

Contact this candidate