UrBox

Overview of job

ABOUT URBOX

UrBox is Vietnam’s leading digital gifting & loyalty platform that enables businesses to integrate their reward & loyalty programs with an ever-growing network of more than 350 brands & 20,000 accepted online and offline stores in Vietnam.

UrBox has the widest network of clients in loyalty in Vietnam and provides loyalty & reward solutions to many big organizations in the aviation, banking, FMCG, Insurance, Real Estate industries and more.

KEY RESPONSIBILITIES:

• Design & Development: Build AI Agents and Multi-agent workflows to automate business processes (e.g., CS Call Analysis, Jira Task Automation, Knowledge Base Assistants).
• LLM Optimization: Experiment with and select optimal models (GPT-4, Claude, Gemini, or Open-source like Llama/Mistral). Implement advanced RAG (Retrieval-Augmented Generation) techniques for proprietary data processing.
• Production Deployment: Collaborate with Data and Backend teams to package AI services into stable, scalable APIs on cloud infrastructure.
• Prompt Engineering & Fine-tuning: Design complex prompt chains and perform model fine-tuning where necessary to improve accuracy for specialized tasks (e.g., OCR bill scanning, Superset report analysis).
• Monitoring & Evaluation: Develop evaluation frameworks to track AI output quality, ensuring safety, reliability, and precision in a production environment.

• Premium healthcare Insurance
• Entitled to a UrBox gift card, which is topped up every month and used to redeem vouchers via the UrBox platform
• 12 days of leave per year and entitled to a valid labor contract, social insurance
• Young professional working environment with plenty of opportunities to learn and grow and promotion opportunities
• Treated to team-building activities and exciting company trips
• Continuous development of hard and soft skills through work and professional training

• Experience: Minimum 3 years in AI/Machine Learning, with at least 1 year of hands-on experience focused on LLMs.
• Programming Skills: Proficiency in Python and common frameworks (PyTorch/TensorFlow).
• AI Stack: Experience with LangChain, LlamaIndex, or CrewAI/Autogen. Strong understanding of Vector Databases (Pinecone, Milvus, Weaviate, or pgvector).
• System Knowledge: Understanding of Data Pipelines, API integrations, and modern Software Architecture principles.
• Preferred Qualifications:

- Experience in MLOps (deploying and monitoring AI models at scale).

- Familiarity with data security standards (knowledge of PCI-DSS is a significant plus).

- Experience with SQL and relational databases, specifically PostgreSQL.