Backend + AI Engineer

Feather is building AI agents that do real work for enterprises. The Backend + AI Engineer will design and build core runtime systems for autonomous AI agents, enabling them to execute workflows and communicate across various channels.

Responsibilities:

• Design and build the core runtime that powers autonomous AI agents
• Architect systems for reasoning loops, planning, tool use, and memory
• Enable agents to execute multi-step workflows across business systems
• Build systems enabling agents to operate across voice, SMS, chat, and email
• Develop real-time conversation pipelines and turn management
• Handle interruptions, context switching, and long-running dialogues
• Develop agent orchestration layers for async and long-lived tasks
• Build DAG/workflow systems coordinating agent decisions and actions
• Enable outcome-based automation (not just conversations)
• Integrate frontier models into production agent systems
• Build prompt pipelines, evaluation harnesses, and guardrails
• Design reliability layers: fallbacks, retries, human handoffs
• Architect event-driven systems handling millions of agent actions
• Build job queues, schedulers, and execution pipelines
• Optimize latency, throughput, and infrastructure cost

Requirements:

• 3+ years building scalable backend or distributed systems
• Strong experience in Python or TypeScript
• Deep understanding of async processing and event-driven systems
• Experience designing production APIs and service architectures
• Experience working with LLM APIs in production environments
• Familiarity with agent frameworks, reasoning systems, or tool use
• Built systems where AI drives real user or business outcomes
• Comfortable owning infra end-to-end
• Strong debugging and performance optimization skills
• Product-minded — you think in workflows, not endpoints
• Built agent tooling (memory, planning, tool execution)
• Experience with workflow engines or orchestration systems
• Familiarity with real-time communication infra
• Experience with RAG, knowledge bases, or retrieval systems
• Exposure to eval frameworks and agent reliability testing
• Worked on customer ops, sales, or support automation