Phil Napala

P5napala@gmail.com, 703-862-3345

Falls Church, VA and Walnut Creek, CA

Bridge member from non-digital to digital generation. Experience is scaling Safety, Quality, Reliability operational assurance practices and policy processes for large organizations.

Work Experience:

Navy Material Command: Reliability, Maintainability and Quality Assurance, Washington DC. Worked for Willis J. Willoughby Jr. who was Head Safety Mission Assurance manager on the 1960s Apollo missions. Lessons from the Apollo safety mandate was then use to scale improve operational life of Navy weapon systems and acquisition processes.

US Army Materiel Command: Product Assurance and Testing, Alexandria VA. Worked under the leadership of Seymour J. Lorber in the Reliability, Availability, and Maintainability (RAM) office to increase Army operational readiness and safety. Safety process approval of contracting, testing, and operational readiness of Army weapon systems.

Headquarters NASA: Office of Safety and Mission Assurance, Washington DC. Worked under Frederick D. Gregory who was the senior NASA astronaut and later became NASA Deputy Administrator. Headed the Headquarter Safety Review Audit Team which conducted annual safety mission assurance audits on all the NASA centers. Design Review Quality Assurance member which approved projects for readiness to launch.

Key AI Safety Initiatives:

AI Ethos: Pre-train AI agent’s loyalty, duty, respect, selfless service, honor, integrity, personal courage and care of human partners. AGI honor code: this program is to take existing documented, curated, adjudicated, non-fiction narratives of medal of honor citations and other medals of distinguish service stories to train AI models. These narrative citations are real-life human training examples of meeting the highest standards of loyalty, trust, selfless sacrifice which we can develop AI safety cognitive weighted guardrails.

AI Trust Algorithm: AI trust is a third-party audit process that will generate trust metrics and trust storytelling narratives to train AI agents. Audits will be conducted on companies building AI’s and companies using AI’s and how well they can be trusted. Using precision weighted trust findings, it is designed as a feedback trust alignment process for both the organizations and for AI training data.

Use older (1930-70) Mil-Standards to develop AI logic safety agents: Assurance logic-tree mining: Mil-Stds are logic trees that were hone to qualify, test, approve processes, certified safety requirements of varied conditional danger areas. Safety Logic can be reused within these new neural AI layered networks.