Staff AI Platform Engineer: Agent Infrastructure

Harvey, Inc.; San Francisco (CA); Hybrid; USD 231,000 - 340,000; Full time

Job summary

Harvey is building a scalable AI Platform that underpins its agentic products. You will design platform-level systems for context management, session state, and memory, enabling reliable AI delivery across multiple foundation models.

You will own the integration and routing pipelines, and help shape evaluation tooling to measure improvements in AI quality. The role requires leadership in backend engineering for AI-powered workflows, hands-on building of internal SDKs, and a bias toward ownership

Qualifications

• 8+ years of experience building backend systems with AI/ML focus.
• Proven track record shipping multi-model AI systems in production.
• Experience with context management, session state, or memory systems.
• Built internal platforms, SDKs, or shared infrastructure adopted by teams.
• Strong judgment in abstractions and pragmatic shipping of features.
• Interest in agentic AI and infrastructure challenges for reliable autonomous systems.

Responsibilities

• Design and build platform-level abstractions for Harvey’s agentic products.
• Own infrastructure for model integration, routing, and evaluation to select the right foundation model.
• Develop evaluation frameworks and tooling to improve AI quality across teams.
• Collaborate with product engineering, PMs, and design to launch AI products.
• Evaluate and integrate latest advancements in AI and agentic systems as they emerge.

Skills

Backend systems

AI/ML engineering

Platform abstractions

Agent infrastructure

Context management

Evaluator frameworks

Developer experience

Autonomous systems

Job description

Harvey is building a scalable AI Platform that underpins its agentic products. You will design platform-level systems for context management, session state, and memory, enabling reliable AI delivery across multiple foundation models.

You will own the integration and routing pipelines, and help shape evaluation tooling to measure improvements in AI quality. The role requires leadership in backend engineering for AI-powered workflows, hands-on building of internal SDKs, and a bias toward ownership