VSF (VinSmart Future)
AI tóm tắt
Skills bắt buộc
Build and operate scalable backend and AI platforms, with Python as the primary language. The role owns production Retrieval\-Augmented Generation (RAG) and agentic capabilities across APIs, data, infrastructure and user\-facing workflows, while welcoming strong engineers with relevant experience in Go, Java, Node.js or other modern technology stacks.
Develop and maintain backend and AI\-platform services, primarily in Python; use Go, Java, Node.js or other suitable technologies when they improve performance or reliability.
Deliver document ingestion and processing, chunking, embeddings, vector/hybrid search, retrieval and reranking, context management, prompt optimization and quality evaluation.
Build multi\-agent orchestration, MCP, A2A, tool integration, state/task coordination, and integrations with LLM providers such as OpenAI and Anthropic.
Create REST/GraphQL APIs, microservices and event\-driven workflows; integrate SQL, NoSQL, vector databases and third\-party services.
Improve concurrency, asynchronous processing, caching, Kafka\-based messaging, observability and large\-scale data processing.
Apply automated testing, code review, documentation, security practices, CI/CD, Docker/Kubernetes, cloud and infrastructure as code.
Partner with AI/ML, Product, Design and Frontend teams; mentor engineers and evaluate emerging AI technologies.
Bachelor's degree in Computer Science, Software Engineering or a related field, with 5\+ years of backend or full\-stack software development experience.
Strong production Python and clean\-code skills; experience with one or more of Go, Java, Node.js or comparable backend languages; sound knowledge of concurrency, threading and event loops.
Hands\-on experience with FastAPI, Flask or equivalent frameworks; REST/GraphQL, microservices, event\-driven architecture, system design, SOLID principles and design patterns.
Hands\-on delivery of RAG/LLM systems covering document processing, embeddings, semantic/vector/hybrid search, retrieval, context and prompt engineering, LLM\-provider integration, multi\-agent orchestration and MCP/A2A; experience with LangChain, LlamaIndex, AutoGen or CrewAI is valued, and 2\+ years in AI platforms, RAG or agents is preferred.
PostgreSQL/MySQL, Redis/MongoDB, vector stores such as Qdrant, Pinecone, Weaviate or Milvus, plus Kafka and asynchronous processing.
Git, Docker, Kubernetes and CI/CD; AWS/GCP/Azure and Terraform/CloudFormation experience are preferred.
Unit/integration/edge\-case testing, OWASP, rate limiting, logging, token management, monitoring/observability, debugging and performance optimization; strong communication and mentoring. React or Vue.js experience is a plus.