Senior AI Platform Engineer

About the role

Airbyte is developing a runtime infrastructure designed to make AI agents reliable and capable of executing actions across enterprise systems. You will build the orchestration layer that allows agents to reason over business data, retrieve verifiable evidence, and perform tasks securely.

Key facts

What you'll do

• Architect and implement the orchestration layer that translates natural language into reliable system execution.
• Create systems for entity resolution, context assembly, and evidence retrieval.
• Develop reusable Skills that capture domain expertise and complex business workflows.
• Build routing logic to coordinate connectors, tools, deterministic code, and language models.
• Own features from initial runtime design through to the end-user experience.
• Develop frameworks for evidence-backed reasoning, citations, and continuous quality evaluation.
• Implement permission models, freshness validation, and action policies.
• Prototype new agent architectures and iterate based on customer feedback.

Requirements

• 7+ years of professional software engineering experience in production environments.
• Deep expertise in backend architecture and distributed systems.
• Proven experience building AI applications utilizing agents, LLMs, RAG, or MCP.
• Strong fundamentals in system design, concurrency, API development, and testing.
• Experience developing workflow engines, developer platforms, or orchestration systems.
• Ability to balance rapid prototyping with long-term production reliability.
• Strong product intuition and the ability to thrive in ambiguous project phases.
• Excellent written communication skills.

Nice to have

• Experience building AI agent frameworks or evaluation systems.
• Familiarity with orchestration tools like LangGraph, Temporal, or MCP.
• Background in knowledge graphs, vector search, or retrieval systems.
• Experience with data infrastructure including Spark, Kafka, Iceberg, Postgres, or modern data warehouses.

Skills & tools

• LLMs
• RAG
• AI agents
• MCP
• Distributed systems
• Orchestration frameworks
• API design
• System architecture

Practical notes

Benefits include flexible PTO with a 25-day annual target, 16 weeks of paid parental leave, comprehensive medical/dental/vision coverage, a 401(k) plan, and professional development stipends. The San Francisco office provides daily breakfast and lunch. Airbyte is an equal opportunity employer and complies with the San Francisco Fair Chance Ordinance. Agency submissions are not accepted.