Staff Software Engineer, AI Platform
$231k - $340k • Remote • New York • FullTime
Posted 14d ago
Job Location
New York
Remote Work Policy
Fully remote
Employment Type
FullTime
Categories
AI Infrastructure Engineer
About the job
Harvey is transforming legal and professional services by combining frontier agentic AI, an enterprise-grade platform, and deep domain expertise. We are building a generational company at an inflection point, scaling fast and defining a new category. The AI Platform team builds the foundational model layer and agent infrastructure that determines the quality of work our agents deliver. This team is early and has significant opportunities to shape the future of AI in high-stakes domains.
Responsibilities
- Design and build platform-level systems for context management, session state, and memory for agentic products.
- Own infrastructure for integrating, routing, and evaluating foundation models to ensure optimal performance for any given context.
- Build shared evaluation frameworks and tooling to enable teams to systematically measure and improve AI quality.
- Create SDKs, platform primitives, and developer tooling to simplify the process of shipping AI-powered features.
- Partner with product engineering, PMs, and design to launch cutting-edge AI products.
- Evaluate, prototype, and integrate the latest advancements in AI and agentic systems.
Requirements
- 8+ years of experience building backend systems, with at least 1 year focused on AI/ML engineering.
- Proven track record of technical leadership across teams.
- Experience building and shipping multi-model or multi-provider AI systems in production.
- Familiarity with context management, session state, or memory systems in AI or distributed systems.
- Experience building internal platforms, SDKs, or shared infrastructure that have been adopted by other engineering teams.
- Strong judgment regarding abstractions and a pragmatic approach to incremental shipping.
- Excitement about agentic AI and the infrastructure challenges of making autonomous systems reliable.
- A bias toward full ownership and navigating ambiguity.