AI Engineer (LLM/AI Agent Developer/Computer Vision)

Top 3 reasons to join us

• We encourage you reach your full potential
• Apply it and new technology in construction
• Work in your way

Job description

• Track A: LLM / AI Agent Developer - Build conversational agents that can understand user requests, maintain context and memory, retrieve the right knowledge (RAG), and call tools/APIs to automate real business tasks end-to-end.
• Track B: Computer Vision Engineer - Develop and deploy vision pipelines for human detection and real-time tracking, fine-tune models for real camera environments, and optimize inference for accuracy, speed (FPS), and stable production deployment.

• Build intelligent AI agents using LLM APIs (OpenAI GPT, Anthropic Claude, etc.).
• Design conversation flows: context handling, memory, tool/function calling, and multi-step task execution.
• Implement RAG, embeddings, and retrieval pipelines to improve accuracy and reduce hallucinations.
• Integrate agents with internal APIs, databases, and knowledge systems to execute real tasks.
• Deploy, monitor, and improve agent performance in production (latency, cost, quality, safety).

• Build and deploy CV systems for human detection, PPE detection, real-time tracking, and activity monitoring.
• Develop and optimize pipelines for live camera streams (RTSP/HTTP), multi-camera setups, and edge/cloud inference.
• Fine-tune and evaluate models (e.g., YOLO, DETR/RT-DETR, segmentation models) for customer or site environments.
• Implement tracking methods (e.g., ByteTrack/DeepSORT/OC-SORT) and counting/zone logic as needed.
• Improve performance: accuracy, robustness, FPS/latency, and resource usage (GPU/edge devices).
• Deliver production-grade CV services: logging, monitoring, alerting, and model versioning.

• Working address: Ground floor, Anna Building, Quang Trung Software City, Trung My Tay Ward, HCMC
• Holiday: Sat, Sun, National holidays

• 2 rounds of interviews: conducted in-person at Vina Takeuchi Co., Ltd. office in Ho Chi Minh City.
• For candidates living outside HCMC, the first round may be conducted online (details will be discussed during scheduling).

Your skills and experience

• Bachelor’s degree in Computer Science, AI, or related field.
• Preferred graduates from top universities (e.g., University of Technology, National Universities).
• Ability to read and understand technical documentation in English.
• Strong coding skills in Python (or Node.js).

• Experience using LLM APIs (OpenAI / Claude / etc.).
• Hands-on knowledge in:
  • Prompt engineering
  • Context / memory management
  • Tool / function calling
  • RAG (Retrieval-Augmented Generation)

• Experience building CV systems for object detection and/or tracking.
• Solid understanding of:
  • Deep learning fundamentals (CNN/Transformers basics)
  • Model training & fine-tuning workflows
  • Evaluation metrics (precision/recall, mAP, IDF1, etc.)
• Familiar with CV frameworks:
  • PyTorch (preferred) and/or TensorFlow
  • OpenCV, video stream processing

• Experience with agent frameworks: LangChain, LangGraph, LlamaIndex, etc.
• Vector databases: FAISS, Pinecone, Weaviate, Milvus, etc.
• Backend / DevOps: FastAPI, NestJS, Docker, CI/CD.
• MLOps: model registry, experiment tracking, deployment automation.
• Edge deployment: TensorRT, ONNX, OpenVINO, Jetson, Coral, etc.
• Experience with surveillance / jobsite monitoring / attendance systems.

Why you'll love working here