Site Reliability Engineer

This is a fully remote position, open to applicants in Singapore.

📋 Description

• Design and manage cutting-edge adaptive, self-correcting, and multi-hop retrieval pipelines.

• Architect Agentic RAG systems with dynamic retrieval control, query decomposition, iterative retrieve-reflect-refine loops, and collaborative multi-agent retrieval.

• Partner with researchers and engineers to identify and implement innovations driven by model capabilities.

• Focus on context management, long-term memory, subagent and multi-agent architectures, self-evolving agents, and execution of real-world tasks.

• Suggest benchmarks and evaluation methodologies for harness-domain and RAG-domain.

• Build benchmark datasets and establish annotation strategies.

• Measure and enhance agent intelligence across various domains, including retrieval efficiency, latency, groundedness, and success rates in tasks.

• Utilize multi-channel user feedback and real-world task data as valuable research signals.

• Design experiments and datasets to continually enhance agent and retrieval performance in production environments.

⛳️ Requirements

• 2-8+ years of hands-on experience with LLM, RAG, and AI agent systems in a production setting.

• Practical experience in building end-to-end production retrieval pipelines.

• Familiarity with embedding models such as BGE and OpenAI.

• Knowledge of vector stores including Qdrant, Milvus, Pinecone, and Weaviate.

• Experience with hybrid search, reranking models, chunking strategies, text cleaning, and parsing multimodal data.

• Implementation experience with Agentic RAG patterns, such as Self-RAG, Corrective RAG, adaptive retrieval, multi-hop decomposition, and retrieve-reflect-refine loops.

• Hands-on experience with Agent Harness runtimes like Pi Agent, AgentScope 2.0, or similar orchestration frameworks.

• Understanding of session recovery, sandbox isolation, middleware/hook systems, multi-tenant runtime, plan/execute loops, and retrieval-grounded tool calling.

• Deep knowledge of LLM and agent mechanisms, including LLM APIs, KV Cache, Agent Loop, Tool Use, Reasoning, Planning, Skills, MCP, Memory, Subagent, and Multi-Agent.

• Strong understanding of Prompt Engineering and Context Engineering.

• Ability to analyze complex problems from first principles, generate original ideas, and advance research from concept to execution.

• Capability to swiftly translate ideas into functional prototypes with rapid experiment iteration cycles.

• Proficient user of agent products, including coding agents and general-purpose agents.

• Skilled in vibe coding and AI-assisted workflows across diverse languages, frameworks, and domains.

• Demonstrated strong learning velocity in software development.

🏝️ Benefits

• Competitive salary and comprehensive company benefits.

• Flexible work-from-home arrangements (subject to the nature of the business team's work).

• Opportunities for career advancement and ongoing learning.

• Collaborate with top-tier talent in a user-focused global organization with a flat hierarchy.

• Enjoy autonomy within an innovative work environment.

• Equal opportunity employer.

• Diverse workforce.