POISE is an agentic AI capability designed to improve early-stage project discovery and requirements gathering across National Grid Electricity Transmission. 
The project will develop and validate the capability within a controlled subset of National Grid systems and selected Strategic Infrastructure / NGET use cases. POISE will help teams identify prior work, map stakeholders, coordinate engagement, track responses and synthesise reusable insight outputs. 
The project builds on a completed discovery and POC definition phase and will move POISE into experimentation and controlled MVP validation. Progression will be subject to a National Grid go/no-go review based on technical feasibility, security and architecture readiness, user validation, benefit evidence and agreed MVP scope. 
If successful, POISE will generate learning on whether the capability can improve the speed, quality and consistency of early-stage discovery and support future scaling decisions.

Benefits

POISE is expected to deliver benefits by improving the efficiency, consistency and quality of early-stage project discovery and requirements gathering. 
The project will test whether agentic AI can reduce the manual effort involved in locating prior work, identifying relevant stakeholders, coordinating engagement and synthesising early insight. By supporting these activities in a more structured and repeatable way, POISE has the potential to help project teams progress from initial idea to evidence-based scope definition more efficiently. 
The main expected benefits are: 
Improved productivity by reducing repetitive coordination and manual synthesis activity; 
Faster access to relevant organisational knowledge and prior project insight; 
More consistent and traceable discovery outputs; 
More structured stakeholder engagement and response capture; 
Better reuse of learning across future projects; 
Improved quality of early project definition, supporting more effective decision-making and reducing the likelihood of avoidable rework. 
The project is also expected to support wider National Grid digitalisation objectives by demonstrating how AI-enabled workflow orchestration can be applied safely within existing enterprise tools and governance arrangements. This will help inform whether the method can be scaled to other Strategic Infrastructure and NGET project contexts. 
The benefits will be tested during experimentation and MVP validation using defined measures such as time to insight, manual effort reduction, quality and consistency of outputs, stakeholder feedback and adoption potential. Financial benefits have been assessed through the project cost-benefit case, with detailed assumptions retained in the supporting internal CBA material.