Mô tả công việc:
About KiotViet:
KiotViet wishes to transfer a massive power of technology to every business model in Vietnam and Southeast Asia through our simple and effective solutions. KiotViet is committed to creating efficient, convenient products for micro, small and medium-size businesses. Over 15 years, with more than 1500 people who form KiotViet in present, we understand that our members of all time are the foundation of KiotViet nowadays and a launchpad for reaching a better KiotViet in the future. KiotViet is committed to creating a work environment that shares success, shares responsibilities.
We are looking for some
developers to join our product team. Joining us, you will have a chance to build the first marketplace in Vietnam with so many technology challenges. As a developer, you will join a growing Data & AI team with mature collaboration across Data Engineering, Data Science, Product, and Engineering functions to build products to enable an excellent experience
What you will do:
1. Build & ship production AI systems
- Architect, fine-tune, and deploy machine learning / deep learning/ LLMs models for real business use cases - owning the full lifecycle from prototype to production;
- Set quality bars for stability, latency, accuracy, and scalability under VN merchant traffic patterns;
- Define rollback / canary strategies; ensure graceful degradation when models fail.
2. LLM & Agentic AI workflows
- Design and build multi-agent systems using LLM orchestration frameworks (LangGraph, LangChain, or equivalent);
- Implement RAG pipelines, knowledge bases, tool use, and memory for agents that operate across merchant workflows;
- Lead prompt engineering, evaluation frameworks, and continuous improvement loops (Langfuse / custom eval pipelines);
- Build conversational AI (chatbots, voice assistants, sales assistants) that natively understand the Vietnamese commerce context.
3. Computer Vision & Document AI
- Develop and productionize models for OCR, image-to-order parsing, receipt understanding, and product image generation/scanning;
- Tune detection / OCR pipelines for low-quality mobile images typical of merchant uploads.
4. NLP & information extraction
- Build search, text classification, intent detection, and structured information extraction services.
5. MLOps & LLMOps & production reliability
- Build and maintain ML pipelines: data ingestion, feature stores, model registry, versioning, A/B testing;
- Set up monitoring for model drift, hallucination rates, conversion impact, Prompt versioning, eval, cost, and latency. Drive cost optimization for inference (LLM API spend, GPU utilization).
6. Cross-functional collaboration & leadership
- Partner with Product Managers and
Business Analysts to translate VN merchant pain points into AI solutions (not just features);
- Work with Data Engineers on data pipelines, quality evaluation, and feature analysis;
- Review code, mentor junior engineers, and contribute to AI engineering standards;
- Evaluate emerging models, frameworks, and partnerships (open-source agent frameworks, foundation model providers) and recommend adoption.