Building Better AI Programs Together

Guest NotesStephen Dooley works with a diverse team of professionals that are responsible for supporting various data management needs, including the discovery process. He has over 25 years of experience advising attorneys, clients, and colleagues on a gamut of complex e-discovery and technical topics, including data governance, data privacy, and GenAI. He tries to remain a resource and thought leader to practitioners throughout the industry through active participation with ILTA.Guest LinkedIn: Stephen DooleyKey TakeawaysDesign for efficiency by eliminating duplication: Look for redundancies and reuse what you build across multiple workflows (e.g., apply lessons from budget summary generation to preservation and legal hold processes).Create a single, transparent source of truth for AI development: A shared program repository keeps everyone educated on what’s new, what people are trying, and—crucially—where things don’t work.Share more to become a stronger team: Competitive instincts can reduce knowledge sharing, but teaching others makes the whole team better.Keep it grounded in real experience—and keep it human: Improve from lived results rather than “we think it can do this,” and remember that tech is “just tech”—the most powerful outcomes come from human purpose and meaning.