About the Role
We're looking for builders who have shipped real AI systems — not just experimented with LLMs. If you've designed production-grade agents that reason over data, make decisions, and trigger real-world actions, we'd love to meet you.
Key Responsibilities
- Design and build production-grade AI agents that reason over business data, make decisions, and trigger real-world actions (planning, inventory, alerts)
- Architect multi-agent systems — planner–executor workflows, tool orchestration, memory, reflection, and retry mechanisms
- Develop agent workflows for retail & supply chain: production planning, store operations, product discovery
- Build RAG & context systems — semantic retrieval and context optimization for LLM reasoning
- Integrate agents with enterprise platforms (Palantir), APIs, and operational systems with guardrails, monitoring, and fallback strategies
Requirements
- 4+ years of experience in Software Engineering, ML Engineering, or AI Engineering.
- Strong Python engineering at production level — not just notebooks
- Hands-on experience with LLMs & Agent systems: advanced prompt engineering, tool/function calling, multi-step reasoning
- Experience building multi-agent workflows with orchestration frameworks like LangGraph or LangChain
- Experience with RAG, embeddings, and semantic search
- Ability to design end-to-end AI systems: data → reasoning → decision → action
- Experience integrating with APIs, microservices, and real-world systems
Nice to have
- Experience with Palantir Foundry / AIP / Ontology or similar platforms
- Background in retail, supply chain, logistics, or operations research
- Experience deploying AI to production with monitoring, evaluation pipelines, and performance optimization
Why Join Us
- Founding team: You help define the architecture and technical direction — not inherit it.
- Real scale: Work on AI systems used across retail, supply chain, and consumer platforms with millions of daily users.
- Rich data: Access large-scale transaction, behavioral, and operational data across Masan's ecosystem — one of the richest enterprise datasets in Vietnam.
- Meaningful problems: Production AI that drives business outcomes, not proof-of-concepts sitting in a drawer.
- Competitive package: Compensation designed for experienced technology talent, with room to grow as the team scales.
