Project design tips and examples: AI chatbots, backlog review and realistic scope

At the webinar, presenters urged applicants to design projects that meet unmet needs and are feasible within a two‑year fellowship. Lauren Hartley recommended collaborative design between the host and candidate, attention to sustainability and realistic impact estimates (for example, whether a project will help 10 or 200 people over two years).

Hartley gave concrete examples of recently funded projects: one fellow will work with Legal Aid of North Carolina to support an “AI‑powered chatbot that helps pro se folks work through the benefits and forms” after Hurricane Helene, acting as the human in the loop to maintain resources and improve the tool. Another fellow will work with the Ohio Innocence Project to apply AI tools to review a substantial backlog of case requests; Hartley described the backlog as “like a 10 or 12,000 request backlog that they have.” These examples illustrate the program’s openness to tech‑assisted projects and to work that pairs legal help with scalable tools.

Presenters cautioned applicants against overambitious litigation plans and recommended projects be distinct from standard staff‑attorney roles, ambitious yet achievable, and designed with a view toward what the project will leave behind (trainings, handbooks, clinics) to increase sustainability and sponsor interest.