Senior AI Engineer, Generative AI Platforms

Your Role

The Senior Staff AI Engineer, Generative AI Platforms & Workflows will lead the engineering of AI-native platforms and automation systems that power content production, creative operations, and cross-functional workflows inside FF's Visual Marketing and broader creative organization.

This role sits at the intersection of AI engineering, agent orchestration, and business workflow automation, with the goal of transforming how FF's marketing, content, and operations teams work — turning manual, time‑consuming processes into scalable, AI‑powered systems. You will design and ship AI agents, tool integrations, and production‑grade automation that enable a small team to deliver with the output of a much larger one.

This is a hands‑on engineering role. You will architect and build the systems yourself — multi‑agent orchestration, memory and knowledge layers, model routing, tool integration, and the user‑facing interfaces that sit on top. You will also contribute to FF's broader EAI strategy by helping develop internal AI platforms that the Company plans to roll out across its marketing, creative, and operations functions.

This position reports to the Head of Visual Marketing and works cross‑functionally with marketing, content, product, and engineering teams to align AI‑powered platforms with business objectives and measurable outcomes.

This is a full‑time on‑site role based in Los Angeles, California, requiring regular in‑office collaboration. Fluency in both English and Chinese is required.

• AI Workflow Engineering: Build AI‑powered workflow systems that help marketing, content, and operations teams automate complex creative and business tasks — from asset generation and campaign operations to research, review, and distribution pipelines.
• Agent Platform Development: Design and implement multi‑agent orchestration infrastructure — state machines, supervisor‑executor patterns, task scheduling, memory and knowledge systems, and failure recovery — so that non‑technical teams can configure and run their own AI teams.
• Generative Content Pipelines: Architect end‑to‑end generative AI pipelines spanning image, video, copy, and multimodal asset generation to support FF's global brand marketing, product launches, and visual storytelling initiatives.
• Tooling & Integration Layer: Integrate AI tools, MCP‑based plugins, and third‑party services (Open Router, Composio, Claude Code, Cursor, Midjourney, Runway, Comfy
  
  UI, and similar) into production‑grade workflows that can be safely operated by business users.
• AI‑Native Engineering Practices: Use AI coding tools (Claude Code, Cursor, and equivalents) as your primary development method. Establish engineering workflows, prompts, and patterns that maximize one‑engineer throughput through AI‑assisted development.
• Business‑Aligned Delivery: Work directly with marketing, content, product, and operations stakeholders to understand pain points, translate messy business requirements into clean technical solutions, and measure impact on time‑to‑ship, content quality, and operational cost.
• Platform & SOP Thinking: Design for reuse and scale — build shared components, asset libraries, SOPs, and self‑serve tooling so creative and business teams can operate AI systems independently, without engineering bottlenecks.
• Innovation & Experimentation: Continuously evaluate emerging models, agent frameworks, and creative AI tools. Prototype rapidly, validate against real business workflows, and bring proven technology into production.

• Education: Bachelor's degree or above in Computer Science, Electrical Engineering, or a related technical discipline.
• 7–10 years of software engineering experience, with meaningful hands‑on work in AI/ML systems, AI agents, or AI‑enabled tools and products.
• AI Engineering: Demonstrated ability to design and ship AI‑powered products — not just calling APIs, but building orchestration logic, prompt systems, memory layers, and tool‑integration chains.
• LLM Fundamentals:
  Solid understanding of LLM capabilities and boundaries — context windows, prompting patterns (System Prompt, CoT, Few‑shot), model selection, and cost/quality…