Trivita AI is hiring a Senior Agentic AI Engineer to build production AI agents, Graph RAG, knowledge graphs, retrieval systems, and evaluation frameworks.
Job description
- Design and build production-grade AI agents capable of multi-step reasoning, planning, tool use, and workflow execution.
- Architect knowledge-intensive AI systems using Knowledge Graphs, Graph RAG, and hybrid retrieval.
- Develop search and retrieval pipelines, including indexing, query understanding, ranking, reranking, and context construction.
- Build evaluation frameworks and benchmarks to measure agent success, retrieval quality, latency, and cost.
- Run rigorous experiments, analyze failure cases, and systematically improve end-to-end agent performance.
- Research and implement relevant techniques from recent papers, validating their effectiveness through controlled experiments.
- Develop robust agent architectures with state management, memory, tool orchestration, and failure recovery.
- Collaborate with backend and infrastructure engineers to deploy, monitor, and continuously improve AI systems in production.
- Own complex technical problems from research and architecture through implementation, evaluation, and production rollout.
Requirements
- Strong hands-on experience building and deploying AI agents or advanced LLM-powered systems beyond basic API wrappers and prompt chains.
- Deep practical understanding of information retrieval, search systems, or knowledge-intensive AI architectures.
- Experience designing evaluation frameworks, building benchmarks, analyzing failure cases, and iterating based on measurable results.
- Strong Python engineering skills and the ability to deliver maintainable, production-ready AI systems.
- Ability to read research papers critically, reproduce relevant methods, and validate whether they deliver measurable improvements.
- Experience making architectural trade-offs involving agent reliability, reasoning quality, latency, and inference cost.
- Strong ownership and problem-solving skills, with the ability to drive ambiguous AI challenges independently.
Technical skills
- Agentic Systems: Tool calling, planning, multi-step execution, agent orchestration, state management, and workflow reliability.
- Retrieval & Search: Hybrid search, dense and sparse retrieval, reranking, query expansion, and retrieval evaluation.
- Knowledge Systems: Knowledge Graphs, Graph RAG, entity resolution, graph traversal, and structured context retrieval.
- Evaluation: Task-specific benchmarks, end-to-end agent evaluation, tracing, error analysis, and systematic experimentation.
- Agent Memory: Short- and long-term memory, context management, memory retrieval, and knowledge updating.
- Production AI: Latency and cost optimization, observability, failure recovery, concurrency, and scalable serving.
Preferred qualifications
- Experience building agents for specialized domains such as Medicine, Law, or other knowledge-intensive industries.
- Experience building evaluation frameworks or benchmark datasets from scratch.
- Familiarity with LangGraph, LlamaIndex, DSPy, DeepEval, Ragas, or equivalent technologies.
- Experience with Neo4j, Qdrant, Elasticsearch, or similar knowledge and retrieval infrastructure.
- Experience with multi-agent systems, agent memory architectures, or reinforcement learning for agent optimization.
- Research publications, open-source AI contributions, or demonstrated technical work beyond standard framework usage.
Contact information
- Address: No. 01, Street 104, Quarter 3, Binh Trung Ward, Ho Chi Minh City
- Phone: [protected info]
- Email: [protected info] / [protected info]