Member of Technical Staff, Applied AI
Finch Legal · New York City
About this role
## About Finch Finch’s mission is to make justice radically more accessible to every American household. We combine expert operators with purpose-built AI to automate the admin work of consumer law—handling intake, claim opening, medical records, police reports, lien management, demands, and more. ## A note from Ben Hey, I’m Ben (CTO at Finch). I’m hiring engineers who can: - Understand what needs to be built and why - Ship quality work fast - Own outcomes tied to measurable legal impact ## What you’ll do (Member of Technical Staff, Applied AI) You’ll build agents that move personal injury cases forward and make workflows dependable. You’ll own critical features end-to-end and work closely with strategy & operations, product, design, and legal. **Responsibilities include:** - Build production-grade **voice, browser, document, and workflow agents** - Design **long-horizon agent systems** (context, tools, memory, permissions, structured outputs, human-in-the-loop controls) - Own the **eval flywheel**: turn real cases and operator feedback into benchmarks to continuously improve agent quality - Optimize performance via **model selection/routing, context engineering, prompting, tool design, and observability** - Make agents reliable across calls, documents, communications, and third-party portals (including failures and changing case info) - Evaluate advances in agentic AI and rapidly ship improvements to **quality, speed, or cost** ## What will make you successful You’ll likely thrive if you have: - **2+ years** building reliable, secure production software; ability to own systems across the stack - Experience shipping **AI-powered products/agents beyond prototypes** - Practical experience designing **evals** and using them to improve task-completion quality - Proficiency in **Python** and comfort with **React, LLM APIs, and agent frameworks** - Strong instincts for agent design: context, model selection, tool boundaries, structured outputs, failure recovery, and human handoffs - A track record of turning ambiguous workflows into simple, dependable systems - An experimental mindset: move quickly, measure what matters, and prioritize correct completion over cleverness ## It’s a plus if you have - Experience with **Pydantic AI** or another production agent framework - Experience building **voice agents, browser agents, or long-running workflow agents** - Familiarity with **Django, React, Temporal, Playwright, AWS** - Experience with **evaluation systems, agent observability, model routing, or HIL review tools** - Familiarity with **legal tech, healthcare workflows, document processing**, or other operationally complex domains ## This role may not be for you if - You prefer a role where roadmap/process/success metrics are already fully defined - You want a narrow remit instead of meaningful ownership on a fast-moving team - You’re looking for a remote-first role (Finch works together in the NYC office **four days a week**) ## What happens next Finch aims for a human, efficient process: 1. **Intro**: conversation about background, motivation, and role 2. **Technical interview**: practical technical session 3. **Onsite**: visit NYC office to interview, meet the team, and have lunch
Listing freshness
CronJobs last confirmed this listing 2h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.