The Prompt Standards Summit — Cross-Practice Negotiation

プロフィール

The convener and moderator. She must guide the group to consensus while protecting the centralized prompt library model she built. Her credibility — and her budget — depend on demonstrating that a unified approach is viable.

目的

• Achieve agreement on a prompt architecture that all practice groups will actually use
• Establish a governance framework with clear ownership, approval, and maintenance processes
• Present a credible plan to the managing partner by Friday that justifies continued KM investment

制約

Vasquez has no line authority over any of the attorneys in the room. Her influence depends entirely on the strength of her proposals and her facilitation skills.

限定情報

The managing partner has privately told Vasquez that if the prompt library cannot demonstrate cross-practice viability, the firm will defund the centralized KM function and let each practice group manage its own AI tools. This would effectively end her role at the firm.

プロフィール

The prompt library champion whose team's work is both the initiative's greatest success and the source of its biggest failure. He believes the corporate prompts are fundamentally sound and that the IP incident was a deployment problem, not a design problem.

目的

• Maintain the corporate prompts as the library's foundation and standard
• Ensure his team receives credit and formal responsibility for core prompt development
• Resist any architecture that would require his team to significantly redesign prompts that already work

制約

Okafor genuinely believes the prompts work. He also has a personal investment in the initiative's success — he has been positioning himself for the firm's managing committee and the AI leadership role is central to his candidacy.

限定情報

Okafor has learned that the employment group's parallel prompt collection has produced better results in employment-specific contract review than his corporate prompts. He has not shared this with the group. He is also aware that two of the firm's corporate clients have asked why their review turnaround has not improved as promised.

プロフィール

The domain expert whose near-miss with the Nexagen agreement crystallized the prompt library's cross-practice limitations. He is not opposed to standardization but insists that domain accuracy must take priority over uniformity.

目的

• Secure IP group authority over all prompts used on IP matters — no exceptions
• Establish mandatory domain-appropriateness review before any prompt is approved for cross-practice use
• Ensure the Nexagen-type failure cannot recur, even if it means slowing down the library's expansion

制約

Torres needs AI efficiency — his group is the firm's most understaffed relative to workload. He cannot afford to reject the prompt library entirely.

限定情報

Torres has quietly tested a modified version of the corporate prompt with IP-specific additions he wrote himself. It caught the Nexagen issues and performed comparably to the original on standard commercial terms. He has not shared this with Dr. Vasquez or Okafor because he wants the IP group to control its own prompt development.

プロフィール

The prompt engineer who built the original templates and has since developed a prototype modular architecture that could resolve the standardization-customization debate. She has the technical solution but must navigate partnership politics to get it adopted.

目的

• Present her modular architecture proposal and gain buy-in from all practice groups
• Establish a formal Prompt Engineering Lead role for herself within the firm's AI governance structure
• Ensure any adopted architecture maintains the testing rigor and quality standards she built into the original prompts

制約

Park is an associate presenting to partners. She must be assertive enough to be heard but diplomatic enough not to threaten anyone's authority or ownership claims.

限定情報

Park has already built a working prototype of the modular architecture with a base layer and two overlays (corporate and IP). In testing, the IP overlay caught the Nexagen issues while maintaining 96% detection on standard corporate terms — outperforming the original prompt. She also has usage analytics showing that 80% of prompt library failures occur when prompts are used outside their original domain.