About this role
Role Overview
Join an innovative initiative aimed at building realistic enterprise environments for training and evaluating frontier AI agents. This role invites experienced public-interest attorneys from major legal-aid organizations, public-interest law firms, and government offices to recreate their digital workspaces and design challenging tasks that reflect real-world legal scenarios.
Key Responsibilities- Construct a realistic digital workspace that mirrors your daily operations, including intake notes, case summaries, pleadings, advocacy letters, policy briefs, grant reports, amicus drafts, and email threads, along with relevant platforms like Westlaw, LexisNexis, Clio, MyCase, Socrata, and ArcGIS Hub.
- Design multi-step tasks based on your actual workflows that require navigating various applications, files, and stakeholders, effectively challenging advanced AI agents.
- Collaborate with fellow public-interest attorneys to shape the environment, define task scope, and review scenarios for realism and rigor.
- Work asynchronously with research teams to refine task designs and establish evaluation criteria for public-interest-law agent benchmarks.
- Contribute to cutting-edge AI research and benchmarking, with your work directly influencing how leading labs train and evaluate future AI systems.
- Juris Doctor (JD) with active bar admission.
- Minimum of 3 years of full-time public-interest experience at a major legal-aid organization, public-interest law firm, state AG/DOJ office, or impact-litigation nonprofit.
- Expertise in areas such as civil rights, civil legal aid, environmental law, consumer protection, impact litigation, or administrative advocacy.
- Proficient in using Westlaw, LexisNexis, Clio, MyCase, and Socrata/Esri ArcGIS Hub in daily practice.
- Strong analytical and writing skills, capable of translating public-interest workflows into structured task specifications.
This project will initially offer an effective hourly rate, transitioning to a compensation model based on the quality and throughput of work produced.