Our AI and Machine Learning teams build systems that help legal professionals reason over case data, understand documents, generate work products, retrieve grounded answers, and automate high-value legal workflows. This work requires strong technical execution, but also deep legal judgment. We need people who can help define what high-quality legal AI output looks like and ensure our systems work in the context of real legal practice.
We’re looking for a Legal Engineer to work closely with our Machine Learning Engineers, Product, Software Engineering, and Data Annotation teams. This role sits at the intersection of legal practice, AI product development, data quality, and model evaluation. It is not a traditional legal counsel role; we want someone who can translate legal workflows into clear product and ML requirements, evaluate model behavior, identify legal edge cases, and help build AI systems that legal professionals can trust.
Key responsibilities
Work closely with ML engineers, data scientists, product managers, and engineering teams to improve AI capabilities across Filevine’s legal workflows
Translate legal tasks, documents, workflows, and user expectations into clear requirements for AI systems, prompts, retrieval pipelines, evaluations, and product behavior
Define what “good” looks like for legal AI outputs, including accuracy, completeness, grounding, citations, tone, usability, and appropriate handling of uncertainty
Review and evaluate model outputs across use cases such as case summarization, document analysis, deposition workflows, demand generation, medical chronology, intake, task automation, and matter-level Q&A
Create and maintain evaluation rubrics, golden answers, annotation guidelines, failure-mode taxonomies, and workflow-specific quality standards
Partner with data annotation teams to design review workflows, calibrate reviewers, resolve ambiguity, and improve the legal relevance and consistency of labeled data
Identify legal and workflow-specific edge cases, hallucination patterns, citation failures, missing context, and product behaviors that could reduce user trust
Provide structured feedback to ML and Product teams on model performance, product gaps, data quality issues, workflow friction, and opportunities for improvement
Help refine prompts, retrieval strategies, agent workflows, and human-in-the-loop review patterns for legal use cases
Who are we looking for
Strong communication and writing skills in English
JD or equivalent legal training, with practical experience in a law firm, legal department, government legal office, or legal operations environment
Strong understanding of legal workflows, legal documents, matter management, and how attorneys, paralegals, and legal staff work in practice
Ability to evaluate legal content carefully, identify missing context or inaccuracies, and explain why an AI output is or is not useful to a legal professional
Strong interest in AI, machine learning, legal technology, document intelligence, and workflow automation
Experience with AI tools, prompt engineering, evaluation workflows, document review, knowledge systems, or data annotation processes is a plus
Ability to create clear rubrics, examples, test cases, review guidelines, and structured feedback for technical and non-technical audiences
Comfort working closely with engineers, ML teams, product managers, designers, and data annotation partners in a fast-moving product development environment
Experience in litigation, personal injury, insurance defense, immigration, mass torts, corporate legal, or other high-volume legal workflows is a plus
Ability to meet periodically in our office, as we value in-person collaboration