About the role
Who you are
- If you thrive in ambiguity, push for excellence, and want to help shape the future of work alongside others who raise the bar, we invite you to build with us
- 5+ years of experience building backend systems, with at least 1+ year focused on AI/ML engineering. Staff candidates will typically have 8+ years and a 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. You’ve thought about what the model sees and why it matters
- A track record of building internal platforms, SDKs, or shared infrastructure that other engineering teams actually adopted - and an understanding of why developer experience matters as much as raw capability
- Strong judgment about abstractions. Opinionated about good design but pragmatic about shipping incrementally
- Excitement about agentic AI and the infrastructure challenges of making autonomous systems reliable when the stakes are real
- A bias toward full ownership: you navigate ambiguity well and don’t wait for a roadmap to start solving problems
- Bonus: experience building evaluation frameworks, working with agent/function-calling architectures, familiarity with legal or other high-stakes professional services domains, or time at early-stage or hyper-growth startups where the underlying technology changes regularly
What the job involves
- Harvey’s products all depend on a shared AI foundation: the model layer and agent infrastructure that determine the quality of work our agents deliver. Legal is one of the hardest domains for AI: documents run to hundreds of pages, matters can span millions of files, and there is zero margin for error on accuracy
- The AI Platform team builds the foundation that every product and agent team at Harvey builds upon. This team is early and there’s a lot to build: model routing, agent architecture, context management, evals. Your work here sets the ceiling for what Harvey’s AI can do
- Context Engineering & Agent Infrastructure. Build the platform-level systems for context management, session state, and memory that all of Harvey’s agents and products rely on
- Model Integration & Routing. Own the infrastructure that lets Harvey onboard new foundation models fast and route to the right one for every task - a capability every product team depends on
- Evaluation Infrastructure. Build the shared eval tooling and frameworks that let every team across Harvey measure and improve AI quality systematically
- Shared Abstractions. Create the SDKs, platform primitives, and developer tooling that make it dramatically easier for product teams to ship AI-powered features
- Design and build abstractions and platform-level systems that improve all of Harvey’s agentic products
- Own infrastructure for model integration, routing, and evaluation that helps Harvey choose and deploy the right foundation model for any given context
- Build evaluation frameworks and tooling that let every team across Harvey iterate on AI quality effectively
- Partner closely with product engineering teams, PMs, and design to launch cutting-edge AI products
- Evaluate, prototype, and integrate the latest advancements in AI and agentic systems as they emerge
Similar jobs you might like
About the role
Who you are
- If you thrive in ambiguity, push for excellence, and want to help shape the future of work alongside others who raise the bar, we invite you to build with us
- 5+ years of experience building backend systems, with at least 1+ year focused on AI/ML engineering. Staff candidates will typically have 8+ years and a 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. You’ve thought about what the model sees and why it matters
- A track record of building internal platforms, SDKs, or shared infrastructure that other engineering teams actually adopted - and an understanding of why developer experience matters as much as raw capability
- Strong judgment about abstractions. Opinionated about good design but pragmatic about shipping incrementally
- Excitement about agentic AI and the infrastructure challenges of making autonomous systems reliable when the stakes are real
- A bias toward full ownership: you navigate ambiguity well and don’t wait for a roadmap to start solving problems
- Bonus: experience building evaluation frameworks, working with agent/function-calling architectures, familiarity with legal or other high-stakes professional services domains, or time at early-stage or hyper-growth startups where the underlying technology changes regularly
What the job involves
- Harvey’s products all depend on a shared AI foundation: the model layer and agent infrastructure that determine the quality of work our agents deliver. Legal is one of the hardest domains for AI: documents run to hundreds of pages, matters can span millions of files, and there is zero margin for error on accuracy
- The AI Platform team builds the foundation that every product and agent team at Harvey builds upon. This team is early and there’s a lot to build: model routing, agent architecture, context management, evals. Your work here sets the ceiling for what Harvey’s AI can do
- Context Engineering & Agent Infrastructure. Build the platform-level systems for context management, session state, and memory that all of Harvey’s agents and products rely on
- Model Integration & Routing. Own the infrastructure that lets Harvey onboard new foundation models fast and route to the right one for every task - a capability every product team depends on
- Evaluation Infrastructure. Build the shared eval tooling and frameworks that let every team across Harvey measure and improve AI quality systematically
- Shared Abstractions. Create the SDKs, platform primitives, and developer tooling that make it dramatically easier for product teams to ship AI-powered features
- Design and build abstractions and platform-level systems that improve all of Harvey’s agentic products
- Own infrastructure for model integration, routing, and evaluation that helps Harvey choose and deploy the right foundation model for any given context
- Build evaluation frameworks and tooling that let every team across Harvey iterate on AI quality effectively
- Partner closely with product engineering teams, PMs, and design to launch cutting-edge AI products
- Evaluate, prototype, and integrate the latest advancements in AI and agentic systems as they emerge