The Centre for Disaster Protection (CDP), funded by UK Aid through the Foreign, Commonwealth & Development Office (FCDO), helps countries and institutions plan and pay for disasters before they strike. As its advisory caseload grew, CDP’s small team faced mounting pressure to digest dense financial documentation, multilingual sources, and a fragmented knowledge base — with ad-hoc ChatGPT use unable to meet the confidentiality and analytical rigor, that the mission demands.
CDP set out to move beyond basic generative-AI usage toward a structured, responsible, adoption plan that is grounded in those values, and aligned with FCDO requirements. Relief Applications partnered with CDP, through its managing agent DAI, on an intensive six-week engagement. We diagnosed workflows, data sources and infrastructure; compared three architecture pathways with cost-benefit analysis; co-designed governance principles with senior leadership, including the Executive Director and the Chief Finance & Operations Officer; and shaped a phased roadmap that explicitly anticipates Phase 2 scope.
Three integrated artifacts — an AI Readiness & Strategic Options Report, an AI Project Governance Strategy, and a High-Level Roadmap — alongside an executive-ready governance framework visual and a leadership presentation. Every document was built to be portable through the upcoming managing-agent transition, values-led rather than compliance-heavy, and immediately usable for senior decision-making.
Recent Comments