Description
This role sits where Harvey’s AI research becomes something lawyers actually use.
The Product Engineering organization builds both the visible product surfaces and the agentic systems behind them. That can mean moving between a React interface, Python backend services, Postgres data models, streaming APIs, long-running jobs, and an agent loop that needs to call tools or operate across multiple steps.
Harvey’s users interact through web applications as well as embedded experiences in tools such as Word, Outlook, and Slack. Product engineers therefore need to think about user experience and reliability at the same time: AI functionality cannot feel like an isolated research prototype inside software used for professional legal work.
At Staff or Senior Staff level, the scope extends across several teams and product lines. Engineers may define shared architecture, establish patterns for long-running AI workflows, design systems used by multiple products, and mentor other senior engineers while still contributing directly to production code.
Agent experience is particularly valuable. Harvey is looking for people who understand harnesses and loops, tool calling, memory, evaluation, and the practical failure modes of autonomous systems—not just developers who have called an LLM API from a backend endpoint.
The broader stack includes React, TypeScript, TailwindCSS, Python, relational databases, backend APIs, and infrastructure supporting streaming or asynchronous processes. Strong engineers from backend or full-stack backgrounds can fit even if agents are not their only area of expertise.
AI-assisted development is also part of the working style. Tools such as Claude Code, Cursor, and Codex are expected to accelerate implementation, but engineers remain responsible for architecture, correctness, maintainability, and product judgment.
Someone comfortable owning ambiguous product problems from architecture through customer-facing delivery would fit this role better than an engineer seeking a narrowly scoped component to maintain.
Curated opportunity. Please verify details and apply via the original link below. No Signals are required for this project/job.
https://www.harvey.ai/company/careers/7797d468-fabc-4bda-b861-726db7171f23