Advisory
An AI knowledge layer for a twelve-thousand-practitioner practice
12,000+
Practitioners with governed AI access
Challenge
A global strategy and transactions practice held decades of proprietary research, sector playbooks, benchmarks and engagement intellectual property across dozens of databases, and none of it was accessible to a generative AI model in a way that respected client confidentiality or IP boundaries. Practitioners were either using public models without the firm’s knowledge behind them, or not using them at all.
Approach
I defined the data architecture and the retrieval-augmented generation framework that sits over the proprietary sources, with confidentiality and IP protection designed into the retrieval layer rather than added afterwards. We set prompt-engineering standards so that outputs were consistent and auditable, and built sector-specific agent kits: pre-configured workflows for common analytical tasks in a given industry, grounded in the firm’s own research rather than the open web.
Outcome
A governed AI layer over the firm’s proprietary knowledge, available to more than twelve thousand practitioners, with sector agent kits that turn a general-purpose model into a domain-specific analyst. The architecture, the standards and the kits gave the practice a defensible way to put generative AI on client work.
Source
Engagement record. Internal programme; no public source.
What decision are you weighing?
For advisory, training and speaking enquiries, or a conversation about a decision you are weighing.