AI Engineer (AI Agent & RAG)

Overview 

How would it feel to be an integral part of the next LinkedIn or SalesForce from an early stage?  Would you like to work with an international team led by Harvard alumni on cutting-edge technology that will transform the future of work around the globe?  Our company recently launched our products to global institutions, and we are continuing to scale.  The best candidates will have opportunity to become key players on our international team that is led by American and Indonesian executives. 

Job Description 

We're looking for an AI Engineer with hands-on experience building agentic AI systems and RAG pipelines. You'll design, build, and maintain the LLM-powered features at the core of our product — multi-step agents, retrieval pipelines, and tool-using assistants integrated into our backend. 

Responsibilities 

• Design and build AI agents with tool/function calling and multi-step orchestration. 
• Build and maintain RAG pipelines end-to-end: chunking, embeddings, hybrid retrieval, reranking, grounded generation. 
• Integrate LLMs into backend services via clean APIs. 
• Own prompt engineering as a discipline — versioned, structured outputs, consistent behavior across providers. 
• Build evaluation harnesses (golden datasets, LLM-as-judge, regression tests). 
• Implement observability for LLM calls — tokens, latency, tool-call traces. 
• Optimize for cost, latency, and reliability: caching, streaming, model routing, retries. 
• Harden agents against prompt injection, tool misuse, and unsafe outputs. 
• Collaborate within backend, frontend, and product teams to integrate AI features.

Requirements 

• 5+ Years as an Engineer
• 2+ years shipping production LLM features (not just prototypes). 
• Strong Python (FastAPI) and/or TypeScript / Node.js for backend AI services. 
• Hands-on with at least one agent framework: LangChain / LangGraph, LlamaIndex, LangSmith, CrewAI, or n8n. 
• Practical experience with native tool/function calling across major providers. 
• Real RAG production experience: chunking trade-offs, embedding selection, hybrid retrieval, reranking, and how to measure retrieval quality. 
• Working knowledge of vector databases: pgvector, Pinecone, Qdrant, Weaviate, or equivalent. 
• Experience designing evaluation pipelines for LLM features. 
• Comfortable with PostgreSQL and RESTful API design. 
• Awareness of prompt injection, PII handling, and safe tool exposure.