
Staff AI Engineer – Agents & Harness
Posted 2 days ago

Posted 2 days ago
This is a fully remote position, open to applicants in Spain.
• Establish the technical vision and create the target architecture for the agent platform, which encompasses execution graphs, LLM interfaces, tool contracts, memory and context management, sub-agents, and sandboxing.
• Compose designs, conduct reviews, and take ownership of architectural trade-offs.
• Advance context engineering through branched conversations, specialist-agent transitions, decision-model layers, and resumable, observable operations.
• Determine agent evaluation criteria utilizing golden question sets, offline and online CI assessments, tracing, feedback loops, and release gates.
• Design the agent-to-agent API, MCP server, headless modes, and exposure boundaries.
• Manage reliability, costs, latency, token budgets, failure responses, guardrails, provenance, and agent security.
• Mentor engineers, guide design and code reviews, and promote agent-building and AI-assisted development methodologies.
• Line-manage Engineers in the Agents team, conduct 1:1s, set objectives, provide feedback, support career progression, and engage in hiring activities.
• Collaborate with the Team Lead, product teams, and customers to define the roadmap, assess build-versus-buy choices, and prototype innovative concepts.
• Remain actively involved with the most challenging technical issues.
• Over 8 years of experience in developing production software, including designing and deploying LLM-powered agents or agentic products utilized by actual users.
• Demonstrated technical leadership without a formal title: established architecture across teams, led design reviews, and mentored peers.
• Extensive, practical understanding of agent systems: tool invocation, multi-agent orchestration, streaming, context and memory management, retrieval, and handling failures over extended operations.
• Strong writing and persuasive skills: articulate trade-offs to engineers, product teams, and executives, and influence decisions with solid evidence.
• Nice to have: developed an agent harness or framework, or contributed to an open-source project (e.g., OpenAI Agents SDK, Claude Agent SDK, LangGraph, DSPy, or similar).
• Nice to have: experience in building MCP servers or clients.
• Nice to have: familiarity with knowledge graphs, ontologies, or entity resolution.
• Nice to have: experience with sandboxed code execution, guardrails, or security measures for agents.
• Nice to have: background in fine-tuning or model selection, or prior research experience.
• Nice to have: interest in commodities or financial markets.
• Proficiency in Python, OpenAI Agents SDK, Postgres, Redis, StarRocks, Kafka, AWS, Nx monorepo, and GitHub.
• Equity in a rapidly growing, VC-backed company.
• $42M in Series B funding — demonstrating strong market traction and investment.
• Opportunities for early-stage growth — advance your career as we expand.
• A market-leading product — we are revolutionizing commodity trading intelligence.
• Access to top-tier tools and technology to facilitate your success.
• Professional development and career acceleration through ownership and impactful contributions.
• Opportunities for mentorship and coaching.
Huron
Amplified Industries, Inc.
HarperCollins Publishers
Get handpicked remote jobs straight to your inbox weekly.