Company Description:
Vin Smart Future is dedicated to improving lives by integrating cutting-edge technology with superior digital experiences. As part of the Vingroup ecosystem, the company aims to build a unified technology platform that seamlessly connects smart operations management with advanced customer applications. Vin Smart Future collaborates with top talent in IT, data science, and AI from both domestic and international communities to create innovative solutions.
Main Responsibilities:
As an Agentics AI Engineer, you will collaborate closely with AI Researchers,
Data Scientists, and
Software Engineers to build autonomous systems capable of reasoning, planning, using external tools, and collaborating with other agents to achieve complex goals. Your responsibilities will include:
Model Development: Developing, and fine-tuning models using state-of-the-art AIalgorithms and techniques mostly for agentic systems, encompassing computervision and NLP.
Research and Experimentation: Experimenting with state-of-the-art AI modelarchitectures in mult-imodal fields and approaches to enhance the performanceand accuracy of existing products. Stay at the forefront of cutting-edgedevelopments in AI and apply them to real-world projects.
Data Preprocessing: Working on data preprocessing, feature engineering, anddata augmentation to optimize and improve model training processes.
Integration: Developing API interfaces for model interaction, closely collaboratingwith software engineering teams to integrate and deploy generalized AI modelsinto real-world applications and systems. This will enhance customer interaction,automate processes, and improve decision-making capabilities.
Job Requirement:
Bachelor's or Master's degree in Computer Science, Machine Learning, ArtificialIntelligence, or a related field.
Experience in building with Multimodal (LLM/VLM or speech) agentic systemsusing Langchain, LLamaIndex or any popular agentic platforms
Proven experience in developing and fine-tuning machine learning models forcomputer vision and natural language processing including Data preparation,model fine-tuning and deployments.
Experience with prompt engineering or context engineering.
Knowledge of Vector Databases (Pinecone, Weaviate, Milvus, or pgvector) forRAG and long-term memory. Building an LLM/VLM model as a core model of the RAG components thatachieved a high ranking on Vietnamese/English datasets is an advantage.
Experience in optimize AI or LLM/VLM models is a plus.
Experience in deployment in serve platform suchs as vllm or Nvidia Dynamo is a plus
DevOps/MLOps: Proficiency with Docker, Kubernetes, and cloud platforms(AWS, GCP, or Azure) is a plus
Evaluations: Experience with frameworks like LangSmith or Arize Phoenix todebug and evaluate agent performance.
Soft skills Requirement:
Systems Thinking: Ability to visualize complex, non-linear workflows and "loops."
Problem-Solving: Strong debugging skills for non-deterministic systems
Ethics & Security: A "safety-first" mindset regarding autonomous actions