P2P

48 THE EXPERTISE-FINDING CAPABILITY: "WHO HAS EXPERIENCE WITH...?" One of the most persistent inefficiencies in any large law firm is finding the right person. When a partner needs to staff a matter involving a niche regulatory issue, or when the business development team needs to identify attorneys with experience in a specific industry for a pitch, the default process has historically been informal: "Pardon the Interruption" (PTI) emails that require lawyers to take time to read and reply if they have the relevant experience, hallway conversations, and slow responses that depend entirely on who happens to see the message and remember the right name. Finding the right information, whether for winning business or delivering on current matters, remains time consuming and resource intensive. This is where AI maturity intersects with data strategy. Expertise finding is often where immature AI breaks down, because it exposes gaps in data quality, governance, and trust. Athena's Experience Connect skill pairs the assistant with curated internal data, including firm intelligence, attorney biography data, and subject matter expertise, to answer questions like "Who has experience with cross-border M&A in the healthcare sector?" The answer is not generated from the model's general training data. It is grounded in the firm's own records, which means the risk of hallucination drops dramatically. The use cases are immediate and practical: faster staffing decisions, stronger pitch and proposal support, quicker responses to client inquiries, and rapidly helping teams find the right expertise. But the strategic value runs deeper. When the assistant can reliably surface institutional knowledge, the firm's collective experience becomes accessible rather than trapped, transforming experience from a static record into a reusable, firmwide capability. OVERCOMING THE SINGLE-CHAT CEILING Most firms began their AI journey with a standalone generative AI chat experience. It drafts, it summarizes, it answers questions. And for a time, that felt transformative. But the novelty faded. The data bears this out. According to Bloomberg Law's 2025 "Analysis: AI in Law Firms: 2024 Predictions; 2025 Perceptions," lawyers were most optimistic in 2024 about AI's potential to automate processes and workflows, with 75% predicting improvement. By 2025, only 37% reported actually seeing that increase. The gap between expectation and experience is the single-chat ceiling in statistical form. This is where many firms remain stuck. The tool works, but it does not integrate into the way attorneys actually practice. It becomes one more application in an already crowded toolbar, useful occasionally but essential never. Breaking through this ceiling requires a fundamentally different architecture that includes agentic capabilities. Moving to agentic architecture also resolved a persistent source of user confusion. The distinction between document analysis skills like "Chat with Document" and "Document Insights," which processed the source document(s)

• Cover